INTERANNUAL AND REGIONAL DIFFERENCES IN KRILL AND FISH PREY

QUALITY ALONG THE WESTERN ANTARCTIC PENINSULA

A Thesis

Presented to

The Faculty of the School of Marine Science

The College of William & Mary in Virginia

In Partial Fulfillment

Of the Requirements for the Degree of

Master of Science

by

Kate Elizabeth Ruck

2012

(!'i/;~ .,.-r

; t ~ •• -~-, _:: I ~-.

APPROVAL SHEET

This thesis is submitted in partial fulfillment of

the requirements for the degree of

Master of Science

Kate E. Ruck

Approved by the Committee, August 2012

~oJA.k.$~6 Deborah K. Steinberg, Ph.D. Committee Chairman/ Advisor

ElizabetJ A. Canuel, Ph.D.

H. Rodger Harfe;, Ph.D //> Department of Ocean, Earth and Atmospheric Sciences Old Dominion University Norfolk, Virginia

CtJ< N~~tf£. Walker 0. Smith, ~hJ)

DEDICATION

For Paul Littreal (1982-20 12)

if miglior fabbro

iii

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS .................................................................... ... v

LIST OF TABLES ................................................................................. vi

LIST OF FIGURES .............................................................................. . vii

ABSTIUCT ....................................................................................... viii

INTRODUCTION ............................................................................................................. 2

MATERIALS AND METHODS ............................................................... .. 6

STUDY AREA AND SAMPLE COLLECTION. .............................................. 6

PREY QUALITY ANALYSES ................................................................... 7 DATA ANALYSIS ......................................................................................................... I 0 DATA VISUALIZATION ............................................................................................. I 0

I{ESUL TS ......................................................................................................................... ll

HYDROGRAPHIC SETTING .................................................................. I I

PREY QUALITY .......................................................................................................... II Spec;es Con1par;son ....................................................................................... II

Sununer of 2009 vs. 201 0 ................................................................................ 13

Reg;onal Colnpar;son. ......... ........................................................................... 14 RelaUonsh;p wUh Length, Laatude, Hydrograph;c Conc/;t;ons, and Thne ..... 15

E. superba Seasonal Compar;son ................................................................... 17

DISCUSSION ................................................................................................................... 17

SPEC'IES C'OMPARISON ........................................................................................... 17 DIFFERENCES IN PREY QUALITY BETWEEN THE SUMMER

OF 2009 vs.2010 ......................................................................................................... 20

REGIONAL COMPARISON OF PREY QUALITY .......................................... 22

RELATIONSHIP OF PREY QUALITY WITH LENGTH, LATITUDE,

HYDROGRAPHIC CONDITIONS, AND TIM£. ......................................................... 25

SUMMARY AND CONCLUSIONS ............................................................................. .28

LITERATURE CITED ................................................................................................... 30

APPENDIX ....................................................................................................................... 58

VITA ................................................................................................................................. 60

iv

ACKNOWLEDGEMENTS

My thesis would never have come to fruition without the support of so many mentors, colleagues, friends and family members. I am most grateful to my advisor, Dr. Debbie Steinberg, for her unflagging energy, humor, guidance, scientific wisdom, and appreciation for the well-executed dance move. She is a remarkable role model for aspiring female scientists and has inspired my respect and admiration from the start. I am so grateful for my admittance into this academic family and the joy for science that it has instilled in me. I am also incredibly thankful for the valuable advice, feedback and assistance provided by my committee members: Drs. Rodger Harvey, Walker Smith, and, especially, Liz Canuel. Liz very generously opened her lab to me, where I completed the bulk of my thesis work, and was an invaluable and enthusiastic expert and endorser of all things lipid.

I also owe a large debt of gratitude to the Principle Investigators and other participants ofthe Palmer Antarctica Long-Term Ecological Research (PAL-LTER) project, especially Drs. Hugh Ducklow, Oscar Schofield, Bill Fraser, Sharon Stammerjohn, Andrew Clarke, and Doug Martinson, and doctoral student Kristen Gorman. Their love and enthusiasm for Antarctic research was infectious and the quality of the work they produce was always a source of motivation for me. Their critical insights and academic support were invaluable to the improvement of my work. Kristen Gorman was especially an inspiration to this project and our collaboration pushed the content beyond its original scope. This project would have also not been possible without the field support provided by Raytheon Polar Services and the captain and crew of the RIV Laurence M Gould, who always went beyond what was expected to ensure we got the samples we needed and made sure much fun was had while getting it all done. Special thanks to Chance Miller, Jamee Johnson, Stian Alesandrini and Mike Coons for making 30 days at sea feel like home.

Many heartfelt thanks to my zooplankton ecology lab members: Joe Cope, Kim Bernard, Lori Price, Brandon Conroy, Josh Stone, Miram Gleiber and Jeanna Hudson. Their support, humor, enthusiasm, and familial love kept me happy and laughing during long hours in the laboratory, in the field, and outside of work. Dr. Grace Saba also deserves special thanks for her companionship, sweetness, and encouragement during the last two years of my thesis. I am also exceedingly grateful for the emotional and technical support of the Canuel lab: Erin Ferer, Paul Littreal (whom we lost too soon), Hadley Mcintosh, Stephanie Salisbury, Matt Mainor and Christie Pondell. Ryan Schlosser was also important in conquering the bomb calorimeter. And general thanks to many other people who have helped with my research: Matthew Erickson, Karen Stamieszkin, Kaycee Coleman, Michael Garzio, Domi Paxton, Kevin Kiley, Jenny Dreyer, Barb Rutan, Andrij Horodysk, Eric Chapman, Eric Lund, Fu-Lin Chu, Quinn Roberts and Heidi Geisz.

Finally, I would like to thank so many other friends and family members who provided sustenance, guidance, critique, and relief during my tenure at VIMS. Thanks to Dan Dutton, Julia Moriarty, Sara Sumoski, Theresa Davenport, Lara Kling, Kelsey Fall, Jaime Blackburn, Allison Watts, Althea Moore, Kellie Nowlin, Gar Secrist, Lea Bennett, Glaucia Fragoso, Kate Rutecki and Scott Marion. And I will always owe the majority of my current and future success to my incredibly supportive and loving family, my parents, Tim and Carolyn, and my siblings, Sarah and Patrick. They have done their utmost to help me achieve my goals and I hope that I will always be a person of whom they can be proud.

v

LIST OF TABLES

Page

Table 1: January 2009 vs. 20 I 0 comparison of environmental parameters and prey

quality .................................................................................................................. 38

Table 2: Regional comparison of environmental parameters and prey quality ................ .41

Table 3: Prey quality measurements by species ............................................................... .42

Table 4: Relationship between lenbrth vs. elemental composition and lipid fraction by

species ....................................................................................................................... 43

Table 5: Relationship between environmental parameters and lipid composition by

species ....................................................................................................................... 45

vi

LIST OF FIGURES

Page

Figure 1: PAL LTER study region and stations sampled .................................................... .46

Figure 2: Regional maps of mean upper water column temperature and

integrated Chl a .................................................................................................. 4 7

Figure 3: Relationship between neutral lipid content and elemental composition by

species ................................................................................................................. 49

Figure 4: January 2009 vs. 2010 comparison of total lipid content by species ................. 51

Figure 5: Relationships between length and total lipid content by species ........................... 52

Figure 6: Regional comparison oftotallipid contents by species ......................................... 53

Figure 7: Relationship between latitude and environmental or temporal parameters ............ 55

Figure 8: Relationship between latitude and total lipid content for Euphausia superba ........ 56

Figure 9: Relationship between length and total lipid content relationship for Pleuragramma

antarcticzun ......................................................................................................... 57

vii

ABSTRACT

Polar zooplankton and fish safeguard against the seasonality of food availability by using the summer months to build large reserves of lipids, which in turn are utilized to meet the metabolic demands of apex predators such as penguins, seals, and whales. A warming trend in the northern part of the western Antarctic Peninsula (WAP) has led to a decrease in perennial and summer sea ice, an increase in heat content over the shelf, and lower phytoplankton biomass, which could affect prey quality. We compared prey quality, including elemental (C, N) content and ratios, total, neutral, and polar lipid content, and energy densities, of known top-predator prey items (krill Euphaush1 superba, 17Jysanoessa macrura, and Euphausia crystallorophias; and fish Pleuragramma antarcticum, and Electrona antarctica) along the W AP latitudinal gradient in January of 2009-20 II as part of the Palmer Antarctica Long-Term Ecological Research study. E. antarctica had the highest prey quality in terms of total lipid content and energy density, followed by T. macrura and P. antarcticum, then E. c1ystallorophias and E. superba. For all species, variations in carbon and nitrogen content were best correlated with by the animals' neutral lipid content, in that animals with larger neutral lipid stores had significantly higher carbon and lower nitrogen content. Across all sexes and maturity stages, E. superba in the South had ca. 20% higher total lipid content than E. superba in the North. Total lipid content was also significantly higher in the South for E. crystallorophias, though this was largely due to the presence of larger individuals in the south combined with a significant positive relationship between length vs. weight-specific total lipid content for this species. For all prey species except T. macrura, there was a positive relationship between latitude or 0-120 m integrated Chi a vs. lipid content (neutral, polar, or total lipids), and a negative relationship between 0-120 m mean water temperature vs. lipid content. Trends opposite to those above found for T. macrura, suggest an optimal habitat for this species in the northern W AP which is characterized by warmer temperatures and lower Chi a. Patterns in Chi a were more important than upper water column temperature in explaining the observed latitudinal trends. If regional warming persists, the prey quality trends described for E. superba, combined with their regional abundance decline in the northern, coastal W AP could affect the ability of apex predators that rely on E. superba to meet their energetics demands.

Kate Ruck

SCHOOL OF MARINE SCIENCE THE COLLEGE OF WILLIAM AND MARY IN VIRGINIA

viii

INTERANNUAL AND REGIONAL DIFFERENCES IN KRILL AND FISH PREY

QUALITY ALONG THE WESTERN ANT ARCTIC PENINSULA

1

INTRODUCTION

The ecosystem dynamics of waters along the continental shelf of the western Antarctic

Peninsula (W AP) are dominated by the seasonal advance and retreat of sea ice which in turn dictates

the development of a strong summer phytoplankton bloom, forming the base of an energy-rich

marine food web (Moline eta!. 2008, Vernet eta!. 2008, Ducklow eta!. 2012a, Steinberg eta!. in

press). Krill are well recognized as an important trophic link in this region, serving as prey for higher

predators such as whales, seals, penguins, and other sea birds (Costa & Crocker 1996, Fraser &

Trivelpiece 1996). Antarctic krill (Euphausia superba) stocks in the W AP are large and persistent

enough to support what some have postulated to be the most important assemblage of endothermic

top predators in the world, in terms of energy flux (Croxall 1992, Atkinson et al. 2004, Ross et a!.

2008).

Recent studies in the W AP also illustrate the importance of fish, especially the endemic,

neritic Antarctic silverfish (Pleuragramma antarcticum) and oceanic myctophids, as a high-energy

supplement to a krill-based diet (Chapman eta!. 2011, Hi1ckstadt eta!. 20 12). The hydrographic

conditions of the WAP are unique in that Upper Circumpolar Deep Water (UCDW), delivered by the

Antarctic Circumpolar Current (ACC), makes its way up onto the continental shelf and becomes

modified UCDW as it mixes with shelfwater (Dinniman eta!. 2012, Martinson & McKee 2012). The

modified UCDW forms a warmer layer in the mesopelagic, allowing oceanic and neritic species of

pelagic fishes to mix, the intensity of which is dependent on the modified UCDW's temperature and

salinity (Donnelly & Torres 2008). The incorporation of these fish species into the diets of animals

classified as highly specialized krill predators can help sustain them through years of low krill

abundance and can be energetically beneficial when provisioning young during key periods of

growth (Chapman eta!. 2011, Hi1cksHidt eta!. 20 12).

2

Climate change is altering the WAP ecosystem and could potentially influence the

availability and nutritional value ofthese prey items for apex predators. The WAP is one of the

fastest warming regions on Earth, with an increase in mid-winter surface atmospheric temperatures

of 6°C since 1950 (Vaughan et al. 2003). The average temperature of the UCDW has also increased

since regular sampling began in 1997, with intrusions of this relatively warm, nutrient-rich water

onto the shelf acting as the main source of heat to the water column (Martinson eta!. 2008). This

regional climate-ocean warming of the WAP has led to major changes in perennial sea-ice dynamics,

characterized by a significant decrease in ice extent and duration (Smith & Stammerjohn 2001,

Stammerjohn et al. 2008, Stammerjohn et al. 2012). As a result, a latitudinal 'climate migration' is

occurring along the Peninsula, with a warmer, more humid, sub-polar climate moving south and

progressively replacing a cold and dry, polar climate (Smith et al. 2003).

This warming of the WAP is affecting the function and structure of the marine pelagic food

web at every level, from primary producers to penguins (Ducklow et al. 2012a). A comparison of two

satellite-derived data sets, spanning nearly 30 years (I 978-1986 and 1998-2006) indicated that

summertime surface Chi a along the W AP declined by approximately 12% between these two time

periods, with a stronger sub-regional decrease (~89%) in Chi a north of 63"S and a substantial

increase (~66%) in Chi a occurring farther south (Montes-Hugo eta!. 2009). These long-term Chi a

trends have larger implications for krill densities and distribution, as E. superba aggregate in areas of

high Chi a (Whitehouse et al. 2008). The abundance of E. superba has also declined in localized

regions of the northern, coastal WAP (Ross et al. 2008, Steinberg et al., unpublished data), mirroring

the larger-scale decline of Antarctic krill in the Southern Ocean (Atkinson et al. 2004). One of the

most dramatic changes in the WAP food web is a roughly 80% decline in breeding pairs of the

native, ice-obligate Adelie penguin in the northern WAP, near Anvers Island, since the mid-1970's

(McClintock et al. 2008, Ducklow et al. 2012a). Sub-Antarctic, ice-intolerant Chinstrap and Gcntoo

3

penguins have in turn established colonies nearby in 1976 and 1994, respectively (McClintock et al.

2008, Ducklow eta!. 2012a).

It has been suggested that decreases in prey quality and abundance could also be contributing

to the decline of Adelie penguins (Chapman eta!. 2010, 2011). Prey quality is mainly dictated by its

energy density, which correlates directly with the animal's lipid content (Clark 1980). Among

searbirds, chick survival and recruitment is often correlated with their mass as they leave the nest

site, and changes in prey q~ality can contribute to a reduction in fledgling mass and, potentially, a

failure in recruitment (Go let et al. 2000, Chapman et al. 20 I 0, 2011 ). This theory is partially

confounded by recent increases of other krill specialists in the WAP, namely Gentoo and Chinstrap

penguins along with humpback and blue whales (Branch eta!. 2007, Ainley et al. 2009, Ducklow et

a!. 2012a). Prey quality and abundances must be high enough to meet energetic demands ofthese

other krill specialists and suggests interspecific competition as another potential variable to explain

the decrease in Adelie penguins and changes in other krill predators (Ainley et al. 2009, Friedlaender

eta!. 2011).

Energetics models are useful in quantifying bottom-up influences of a complex system by

focusing on how external regulators, like the nutritional value of available prey, affect growth.

Recently, an individual-based energetics model was created to simulate growth of Adelie penguin

chicks in colonies along the W AP (Chapman eta!. 2010, 20 II). Their results highlighted the

importance of gravid female E. superba and larger (>70mm) P. antarcticum to the diets of chicks,

where a deficiency of 0.117 kg in fledging mass could result in a chick's failure to recruit. In these

types of energetics studies, the biochemical composition of prey fed to chicks is usually estimated

from length-specific literature values. These estimates do not consider the potential effects that

increasing temperatures and a changing food base could have on energy storage functions for these

mid-trophic level prey species.

4

The metabolic rates of ectotherms vary with ambient temperature and a temperature-driven

increase in metabolic rates could inhibit prey species' ability to accumulate and store lipids in a

region like the northern part of the peninsula, where phytoplankton is also decreasing (Peck et al.

2004, Montes-Hugo et al. 2009). These changes could also impact the metabolism of apex predators.

Digestion efficiency in seabirds decreases as the lipid content of their prey decreases, meaning they

are less effective at assimilating the maximum amount of energy available in lower quality foods

(Brekke & Gabrielsen 1994). Energetics models rarely address parameters besides total lipid content

that contribute to the nutritional value of prey. Jackson & Place (1990) showed that Rockhopper

penguins have higher assimilation efficiencies for triglycerides when compared to wax esters,

showing that different lipid classes can impact penguin metabolism. In addition, phytoplankton with

high carbon to nutrient ratios is of low nutritional value to zooplankton. Consumption of high C:N

phytoplankton can actually shift zooplankton elemental composition, forcing zooplankton into a

nutrient-limited state, as the nutrient content of their food is not sufficient to meet their nutrient

demands (Van de Waal et al. 2009). The caloric content of krill used in models is usually calculated

from other measurements of the animals' biochemical composition. Only three studies to date have

taken direct calorimetry measurements of E. superba (Nagy & Obst 1992, Ainley et al. 2003, Farber­

Lorda et al. 2009), showing that spent females (eggs have been released) have a lower caloric content

than mature males. Incorporating these relationships between prey quality and predator metabolism

into energetics models will increase model accuracy by improving the quality of functions regulating

modeled predator growth.

The climate gradient of the WAP makes it a unique region to address how differing

environmental parameters can affect prey quality, such that prey in the northern W AP may differ in

species composition or biochemical composition, than prey found farther south-where summer sea

ice still persists. To assess whether regional warming and subsequent changes in phytoplankton are

affecting prey quality along theW AP, we compared total, neutral, and polar lipid content, elemental

5

\

(C:N) ratios, and energy densities of known top predator prey items (krill E. superba, 1J1ysanoessa

macrw·a, and Euphausia c1ystallorophias; and fish P. antarcticum and Electrona antarctica)

collected along the W AP latitudinal gradient. Comparing prey quality metrics from different years

and regions with different environmental characteristics allowed us to examine if changes in prey

physiology are related to parameters such as water temperature and Chi a. These relationships will

be useful in future modeling efforts used to predict effects of regional warming on prey quality,

enhancing our understanding of food web energy transfer in this dynamic ecosystem.

MATERIALS AND METHODS

Study area and samrle collection

Zooplankton and fish used for prey quality analyses were collected on the annual, austral

summer cruises of the Palmer Antarctica Long-Term Ecological Research (PAL L TER) program

aboard the A.R.S.V Laurence M Gould during January, 2009-2012. The PAL LTER study,

encompasses 41,000 square nautical miles with regular sampling stations located in coastal,

continental shelf, and oceanic slope zones, and in seasonal sea-ice zones (Ducklow et al. 20 12a).

Based on existing sea-ice dynamics described by Stammerjohn et al. (2008), we divided the study

grid into North (all stations north of, but not including, the 300 sampling line) and South (all other

stations) sub-regions (Figure 1).

Macrozooplankton were collected using a 2x2 m-square-frame net (700 Jlm mesh), towed

obliquely from the surface to 120m, or occasionally shallower in coastal areas (Ross et al. 2008,

Bernard et al. 20 12). The net target depth was determined in real-time with a net-depth sensor housed

in the termination of the tow conducting cable and confirmed with either a Vemco Minilog

Temperature-Depth Recorder or a Star-oddi data storage tag (DAT). The volume of water filtered

through the net was detennined with a General Oceanics flow meter. Once on board, the contents of

the cod end were gently transferred to a large tub filled 'with ambient surface seawater and

6

zooplankton or fish species of interest were removed. We focused on organisms that were previously

identified as important prey items for penguins in colonies along the WAP, including: three krill

species (Euphausia superba, Euphausia clystallorophias, Thysanoessa macrura) and two fish

species (the myctophid Electrona antarctica and the Antarctic Silverfish Pleuragramma

antarcticum) (W. Fraser, pers. comm.). Animal lengths were recorded for E. superba (Standard

length 1, defined as the length between the eyes to the tip of the telson by Mauchline 1970), and fish

(Standard Length, defined as the length between the most forward part of the head and the posterial

end ofthe hypural bone; Total Length, defined as the length between the most forward part of the

head and the end of the caudal fin rays, after Ricker & Westrheim 1979), and sex/maturity stage was

determined for E. superba (Makarov & Denys 1981 ). Total lengths for T macrura and E.

c1ystallorophias were calculated from wet weights based on literature values (Mayzaud et al. 2003,

Ju & Harvey 2004). All animals for prey quality analyses were then frozen and stored at -80°C.

Krill collected in 2009 and 20 I 0 span the entire sampling grid and represent the majority of

the data presented here. Fish were rarely collected, and thus cover a limited geographic and temporal

(one sample from 2009 and the rest in 2010 and 20 II) range. The grid was sampled from north to

south over the course of 5 weeks, thus a seasonal comparison of E. superba was conducted in 20 l 0

and 2012 in which krill were collected from one of the northern-most stations at the beginning and

end ofthe cruise (6 January 2010 & 31 January 2010; 6 January 2012 & 30 January 2012) to

determine if the seasonal progression of summer was potentially biasing observed longitudinal

trends.

Prey quality analvses

In the laboratory, animals were thawed, rinsed three times with Milli-Q water, and placed in

pre-com busted and pre-weighed glass tubes. Animals were either analyzed individually or combined

with other animals ofthe same species and physical characteristic (e.g., size, sex) to form

representative composite samples for individual sampling stations. T macrura composite samples

7

consisted of I 0 individuals, E. c1ystallorophias of 5 individuals, Juvenile E. superba of 2 individuals;

all fish samples were analyzed individually. To test whether the composite samples were artificially

altering variance, we analyzed half of the mature E. superb a as individuals and half as 2-individual

composites. An F-test was then used to compare the variances between these two groups and no

significant differences were found for any of the prey metrics measured. This confirmed that the

variances associated with the composite samples were still representative of individual adult E.

superba in nature. Artificial composites were then created by averaging prey metric values of two

morphometrically similar individual E. superba samples to use in data analysis along with the 2-

individual E. superba composites.

Sample wet weights (WW) were determined with a Sartorius BP211 D analytical balance, and

then homogenized either using a Misonix ultrasonic liquid processor XL-2000 series or a Virtis "45"

homogenizer. The samples were then freeze-dried using a LabconocoFreezone6 Plus freeze dryer and

dry weights (OW) were determined. Sub-samples of the dry, homogenous powder were taken for the

various prey quality analyses. When not in use, samples were stored in airtight containers with

headspace that was flushed with N2 gas to prevent sample degradation.

To determine total organic carbon and nitrogen content, a small sub-sample of freeze-dried,

homogenized tissue was transferred to a pre-weighed, pre-combusted tube, placed in a desiccator

with a small beaker of concentrated HCL for 16 h to remove organic carbonates, and then moved to a

60°C drying oven for a minimum of72 h. The acidified sample was then packed into tin capsules and

the final sample weight was obtained using a Sartorius XP I OOOP microbalance. The sample was then

processed using a Costech ECS 40 I 0 CHNSO Analyzer for flash combustion with acetanilide as the

standard. Roughly I 00 representative krill samples and all fish samples were weighed pre- and post­

acidification using a Sartorius BP2Il D analytical balance to estimate the mass of the organic

carbonates lost to acidification; there was no discern!ble mass difference.

8

E. superba and E. ClJlS/allorophias accumulate both neutral (mainly triglycerides and wax

esters, respectively) and polar (mainly phosphatidylcholine) lipids during the summer as a

mechanism for energy storage for the winter, when food is scarce (Hagen et al. 1996, Ju et al. 2009).

Separately analyzing neutral and polar lipid concentrations enables us to detennine the relative

influence of each in driving any observed total lipid trends. Neutral and polar lipid classes were

extracted and separated utilizing a modified accelerated solvent extraction (ASE) method developed

by Poerschmann and Carlson (2006). Briefly, the sample is placed in a cell with a silica-based

sorbent at its outlet, the cell is pressurized and cycled through two n-hexane/acetone (9: 1, v/v)

extractions at 50°C to collect the neutral lipids followed by two chloroform/methanol (I :4, v/v)

extractions at 80°C to collect the polar lipids. Each fraction was then dried completely under N2 gas

and resuspended in 500f!L of hexane for the neutral fraction and 2000flL of chloroform for the polar

fraction. A known volume of the resuspension was placed into a warm, pre-weighed aluminum cup.

The solvent was allowed to evaporate and the tin cup was reweighed using a Sartorius XP 1 OOOP

microbalance. Three aliquots of each sample were weighed, to calculate a mean and standard

deviation. If the coefficient of variation (standard deviation/mean x 1 00) was less than 10%, the

mean weight was used to calculate total lipid extracted from the sample. lfthe coefficient of variation

was greater than 10%, a fourth aliquot was weighed.

To determine caloric/energy content, a subsample of freeze-dried, homogenized tissue was

formed into a pellet and ignited in a bomb calorimeter (Parr Instrument Co. Model 6300) using

benzoic acid as a standard. The fish samples were comprised of one fish each and supplied enough

dry mass to run a sample replicate. The percent difference was calculated for the replicates to analyze

precision, and the two replicates were averaged to estimate the caloric density (kJ/g dry mass) of the

sample. There was not enough dry mass to run replicates on krill samples, which consisted of a

composite oftwo morphometrically similar individual krill from the same station. The fish sample

9

replicates were always within± 3% of each other; the same precision for the individual krill samples

was assumed.

Data analv.'>is

Statistical differences between prey species, years, and North-South sub-regions sampled

were calculated using a 2-way ANOVA when data met assumptions of normality and homogeneity.

A Two-sample t-test was used to examine data sets that only spanned one year or one region. Data

sets that did not conform to a normal distribution were In-transformed. Data that did not meet the

normality assumption after transformation were tested with the non-parametric Mann-Whitney test.

Regression analyses were used to explore relationships between prey quality metrics and other

environmental and temporal data which were obtained from the PAL L TER database

(pal.lternet.edu/data/). Discrete Chi a values (taken every 15m, on average) were integrated and water

temperature measurements (taken continuously via CTD) were averaged over the top 120m for each

station where prey were collected. Significance for all statistical analyses was determined at a= 0.05.

Data visualization

Maps were created with ArcGIS 10 Geostatistical Analyst to visualize 120 m integrated Chi a

and mean water temperatures for the PAL L TER study region. Temperature interpolations were

created with Ordinary Kriging and Chi a interpolations were created with the Inverse Distance

Weighted (lOW) method. Oceanographic stations from the PAL LTER cruise with available water

temperature and Chi a data were used in the analysis to increase the accuracy of the interpolations

(stations depicted in Figure 2). Regression analyses only utilized data from stations where prey

species were sampled concomitantly (stations depicted in Figure 1 ).

10

RESULTS

Hydrographic Setting

Upper water column (0-120 m) mean water temperatures at stations sampled for prey items

were significantly warmer across the grid in January 2009 than 20 l 0, averaging -0.0 l oc and -

0.58°C, respectively (Figure 2A,B; Table I). For combined 2009 and 2010 data, mean upper water

column temperatures were significantly warmer in the North, especially along the coast, than in the

South, averaging 0.12 oc and -0.57 °C, respectively (Figure 2A,B; Table 2). Upper water column (0-

120m) integrated Chi a was significantly lower in 2009 than 2010, averaging 52.2 mg m-2 and 119.4

mg m·2, respectively (Figure 2C,O; Table 1). Integrated Chi a (2009 and 2010 combined data) was

significantly lower in the North than in the South, averaging 60.8 mg m-2 and 103.8 mg m-2,

respectively with maximum values in the southern coastal regions and Marguerite Bay (Table 2).

There was a wider range in both temperature and Chi a in 20 l 0 compared to 2009, with a difference

in maximum and minimum mean upper water temperature and integrated Chi a of0.26 oc and 86.1

mg m-2 in 2009 compared to those of 1.32 oc and 598.1 mg m·2 in 20 l 0. In 2011, integrated Chi a

averaged 209.2 mg m-2 across the grid, significantly higher than in 2009 (Table 1).

Prey Quality

Spedes Comparison

Average nitrogen content for the three species of krill and P. antarcticum ranged from 0.083

- 0.092 mg N ow-1, and T. macrura and P. antarcticum average nitrogen content was significantly

lower (p<0.05) than E. crystallorophias and All E. superba (Table 3). Average carbon content for

these species ranged from 0.42- 0.49 mg C ow-1, with P. antarcticum carbon content significantly

higher than T. macrura, and both significantly higher than E. ClJISfallorophias and All E. superba

(Table 3). Average C:N ratios by mass of T. macrura and P. antarcticum (5.48 and 5.73,

respectively) were significantly higher compared to E. ClJISfallorophias and All E. superba (4.53 and

11

4.77, respectively) (Table 3). E. superba males had significantly higher nitrogen content and

significantly lower carbon content and C:N values than juveniles and females (Table 3). The

elemental composition of E. antarctica was significantly different from all other prey species, with a

lower nitrogen content (0.059 mg N mg·1 ow-1) and a higher carbon content (0.61 mg C mg·1 OW-1

),

resulting in a the highest C:N ratio (I 0.3) of all species analyzed (Table 3).

The prey species with the highest overall lipid content was E. antarctica (55.0% OW),

followed by T. macrurca (45.0% OW), and P. antarcticum (38.6% OW) (although for the latter two

species, lipid content was statistically indistinguishable). The total lipid content of the two fish

species was mainly composed of neutral lipids, while T. macrura neutral and polar lipid contents

were equal (Table 3). E. crystallorophias and All E. superba had the lowest overall total lipid

contents (29.1% OW and 32.3% OW) (Table 3). For E. superba, males had significantly lower total

lipid content (27.4% OW) than either juveniles or females (34.5% OW and 34.4% OW) (Table 3). E.

c1ystallorophias and every sex/maturity stage of E. superba had higher polar lipid contents than

neutral lipid contents (Table 3).

For all species analyzed, variations in carbon and nitrogen content were best explained by the

animal's neutral lipid content, in that animals with a higher lipid composition were associated with

higher carbon content and lower nitrogen content (Figure 3). The neutral lipid fractions of species in

this study were mainly composed from triglycerides or wax esters. For E. superba and P.

antarcticum, species whose neutral lipid content are mainly composed oftriglycerides (as reviewed

in Falk-Petersen et al. 2000, Lee et al. 2006), neutral lipid content vs. mg C mg-1 ow-1 or mg N mg·1

ow-1 were both best represented by linear regressions (Figure 3). ForT. macrura, E.

c1ystallorophias, and E. antarctica, species whose neutral lipid content was mainly composed of wax

ester (as reviewed in Falk-Petersen et al. 2000, Lee eta!. 2006), neutral lipid content vs. mg C mg· 1

DW-1 was best represented by a linear regression, whjle lipid content vs. mg N mg·1 ow-1 was best

represented by a quadratic polynomial (Figure 3). All regressions were significant (p<0.05)

12

Energy density values for All E. superba, P. antarcticum and E. antarctica were significantly

different from each other. E. antarctica had the highest energy density (31.9 kJ g"1 DW"1), followed

by P. antarcticum (24.6 kJ g-1 DW-1), and then by All E. superba (21.1 kJ g·1 DW-1) (Table 3).

Female E. superba energy density was significantly higher (22.0 kJ g-1 DW-1) than male E. superba

(19.5 kJ g-1 ow-1) (Table 3).

Summer o(2009 vs. 2010

T macrura and E. c1ystallorophias exhibited the most variability between years sampled of

the five prey species investigated. T macrura had a higher average total lipid content in 2009 (54.2%

DW) than 20 I 0 (3I.2% DW), while E. c1ystallorophias showed the opposite trend, with lower total

lipid content in 2009 (2I.O% DW) than 2010 (34.5% DW) (Figure 4, Table I). Prey carbon content

between years followed that of total lipids- higher forT macrura and lower for E. C1JIS/allorophias

in 2009 vs. 20 I 0. The trend was opposite for nitrogen (lower for T macrura and higher for E.

C1J1Stallorophias in 2009 vs. 2010), resulting in higher C:N values in 2009 than 20 I 0 (5.88 and 4.79,

respectively) forT macrura, and lower C:N in 2009 than 2010 (3.92 and 4.94, respectively) for E.

C1J1Stallorophias (Table I). These interannual differences can be partially explained by differences in

average individual total length of animals sampled in 2009 vs. 2010. There is a significant, positive

relationship between average individual total length and total lipid content/carbon content/C:N for

both species (Figure 5, Table 4) and a significant, negative relationship between nitrogen content and

average individual total length for both species (Table 4). The average individual total lengths ofT

macrura sampled in 2009 were significantly higher (14.2 mm) than those sampled in 20IO (12.3 mm)

(Table I). The reverse is true for E. c1ystallorophias, where the average individual total length was

lower in 2009 (29.8 mm) than in 2010 (33.7 mm) (Table I).

Differences between January of2009 and 20 I 0 were less prevalent in the other prey species

of interest. Juvenile and All E. superba had slight but _significantly higher carbon content in 2009

13

than in 2010. The C:N, total lipid, total length, and energy density of the fish P. antarcticum was

higher in 2011 vs. 20 I 0 (Table 1) (there was only one P. antarcticum fish collected in 2009, which

was not included in the analysis). P. antarcticum sampled in 2011 were on average larger ( 111.5

mm) than those sampled in 2010 (91.0 mm), but there were no significant relationships between any

ofthe prey quality metrics of interest and total length (Tables 1, 4). P. antarcticum had significantly

lower average C:N ratios, lipid content and energy density in 2010 vs. 2011 (Figure 4, Table I). (All

of the E. antarctica collected were from 2011, Table I).

Regional Comparison

Species/life stages for which there were significantly higher total lipid content in the South

(all years combined) included E. Cl)JStallorophias, juvenile E. superba, female E. superba, and All E.

superba (Figure 6, Table 2). While not significant, male E. superba and E. antarctica total lipid

content were also generally higher in the South. T. macrura was the only species analyzed with

higher lipid content in the North, though this difference was not significant (Table 2).

Nitrogen content was significantly higher for female E. superba and Electrona sp. in the

North (Table 2). T macrura, E. crystallorophias, male E. superba, and All E. superba also had

higher average nitrogen content in the north, but not significantly so. Carbon content was

significantly higher in the south for T. macrura and All E. superba (Table 2). All other prey

species/life stages (E. ClJiSfallorophias, juvenile E. superba, male E. superba, female E. superba, E.

antarctica) had higher average carbon content in the South, but this trend was not significant. These

elemental trends resulted in higher C:N in the South for ail prey species, with significant higher

South C:N values for T. macrura, E. crystallorophias, female E. superba and All E. superba (Table

2).

The significant North-South difference in E. CIJIS/allorophias total lipid content and C:N

values may again be partiaiiy explained by the sjgnificantly higher average individual length of

14

than in 2010. The C:N, total lipid, total length, and energy density of the fish P. antarcticum was

higher in 2011 vs. 2010 (Table 1) (there was only one P. antarcticum fish collected in 2009, which

was not included in the analysis). P. antarcticum sampled in 2011 were on average larger (111.5

mm) than those sampled in 20 I 0 (91.0 mm), but there were no significant relationships between any

ofthe prey quality metrics of interest and total length (Tables 1, 4). P. antarcticum had significantly

lower average C:N ratios, lipid content and energy density in 2010 vs. 20 II (Figure 4, Table I). (All

of the E. antarctica collected were from 2011, Table l ).

Regional Comparison

Species/life stages for which there were significantly higher total lipid content in the South

(all years combined) included E. crystallorophias, juvenile E. superba, female E. superba, and All E.

superba (Figure 6, Table 2). While not significant, male E. superba and E. antarctica total lipid

content were also generally higher in the South. T macrura was the only species analyzed with

higher lipid content in the North, though this difference was not significant (Table 2).

Nitrogen content was significantly higher for female E. superba and Electrona ::,p. in the

North (Table 2). T macrura, E. cJystallorophias, male E. superba, and All E. superba also had

higher average nitrogen content in the north, but not significantly so. Carbon content was

significantly higher in the south for T macrura and All E. superba (Table 2). All other prey

species/life stages (E. ClJlStallorophias, juvenile E. superba, male E. superba, female E. superba, E.

antarctica) had higher average carbon content in the South, but this trend was not significant. These

elemental trends resulted in higher C:N in the South for all prey species, with significant higher

South C:N values for T macrura, E. ClJlS!allorophias, female E. superba and All E. superba (Table

2).

The significant North-South difference in E. ClJlStallorophias total lipid content and C:N

values may again be partially explained by the sig~ificantly higher average individual length of

14

animals collected in the South (33.1 mm) vs. in the North (28.2 mm) (Table 2). As mentioned above,

total lipid content scales positively with length of E. cJystallorophias (Figure 5, Table 4). So, the

significantly higher total lipid content and C:N values in the South (31.7% DW and 4.72,

respectively) vs. the North (19.0% DW and 3.77, respectively) are more likely due to the different

size distribution of animals in each region (i.e., larger in the south) rather than the impact of differing

environmental parameters on animal physiology (Figure 6, Table 2). Juvenile E. superba total lipid

content is also significantly, positively correlated with length (Figure 5, Table 4). However, juveniles

of various lengths were sampled uniformly across the grid, so the significantly higher total lipid

content in the South (37.6% DW) vs. the North (31.0% DW) is more likely due to regional inf1uences

on animal physiology (Figure 6, Table 2). There were no significant regional differences in energy

densities for any prey species sampled (Table 2).

Relationship vvith length, latitude, hvdrographic conditions, and time

Individual average total length was a good predictor for C:N and total lipid content in T.

macrura (R2 = 0.57 and 0.31, respectively), E. Cl)JStallorophias (R2 = 0.64 and 0.48, respectively),

and to a lesser extent juvenile E. superba (R2 = 0.28 and 0.21, respectively) (Figure 4, Table 4).

Nitrogen content was significantly, negatively related to length for all three species, and carbon

content was significantly, positively related to length for only T. macrura and E. ClJiSfa!lorophias

(Table 4). These individual elemental trends resulted in a significant, positive relationship in length

vs. C:N for all three krill species (Table 4). The neutral lipid fraction is responsible for the majority

of the variability for the significant, positive relationship in length vs. total lipid content for T.

macrura and E. ClJ!Sfallorophias, while both the polar and neutral lipid fractions contribute equally to

the significant positive relationship in length vs. total lipid content for juvenile E. superba (Table 4).

Latitude, mean upper water column temperature, Julian date sampled, and integrated Chi a

are all highly correlated with each other due to the structure of the sampling cruise (starting in the

15

north at Adelaide Island and ending south at Charcot Island) (Figures I, 2, and 7). All relationships

between these properties were significant, with the exception of Julian Date Sampled vs. integrated

Chi a (Appendix lA). Latitude was the most important explanatory variable, resulting in the highest

R2 values for the relationships between mean upper water column temperature (R2 = 0.47), Julian

date (R2 = 0.46), and integrated Chi a (R2 = 0.20) (Figure 6). Mean upper water column temperature

was negatively correlated with latitude, while Julian Date Sampled and integrated Chi a were both

positively correlated with latitude (Figure 7).

T macrura total lipid content was significantly, positively related to mean upper water

column temperature (R2 = 0.39) and significantly, negatively related to integrated Chi a (R2 = 0.79)

(Table 5). The polar lipid fraction accounts for most of the variability in the relationship between

upper water column temperature vs. total lipid content, while both neutral and polar fractions

contribute equally to the relationship between integrated Chi a vs. total lipids (Table 5). E.

crystalloroph;as total lipid content was significantly, positively related to latitude (R2 = 0.35) and

integrated Chi a (R2 = 0.63), and significantly, negatively related to upper water column temperature

(R2 = 0. 70) (Table 5). The neutral lipid fraction accounts for most of the variability in relationships

between environmental parameters and E. crystallorophias total lipid content (Table 5).

All E. superba total lipid content was positively, but weakly (R2 = 0.13 for males and 0.05

for All), related to latitude and integrated Chi a (Figure 8, Table 5), with neutral and polar lipid

fractions contributing equally to the relationship in latitude vs. total lipid content, and the polar lipid

fraction to Chi a vs. total lipid content. Juvenile and male E. superba trends mirrored those found for

All E. superba, with both life stages significantly, positively related to latitude, and both neutral and

polar lipid fractions accounting equally for the variance (Figure 8, Table 5). Male total lipid content

was also significantly, positively related to integrated Chi a, with the polar lipid fraction driving the

relationship (Table 5). Juvenile and female E. superba total lipid content were both significantly,

16

positively related to Julian day and the polar lipid fraction explained the majority of the variance for

juveniles, while the neutral lipid fraction explained more of the variability for females.

There were no significant relationships;between any of the environmental parameters and P.

antarcticum total lipid content, most likely due to the limited regional range of this fish (Figure I,

Table 5). E. antarctica neutral lipid content was significantly, positively related to latitude (R2 =

0.44) and polar lipid content was significantly, negatively related to latitude (R 2 = 0.5 8) (Table 5). E.

antarctica also had a significant, negative relationship between polar lipid content and Julian day (R2

= 0.61).

E. superba seasonal comparison

The grid was sampled from north to south over the course of 5 weeks, thus a seasonal

comparison of E. superba was conducted in 20 I 0 and 2012 in which krill were collected from one of

the nmihern-most stations at the beginning and end ofthe cruise to determine if the seasonal

progression of summer was potentially biasing observed longitudinal trends. There was no significant

difference between total lipid content for male E. superba sampled from the same nmthern station at

the beginning (6 January 2010; 21.6% DW) and end (31 January 2010; 20.3% DW) ofthe cruise in

2010. Male and juvenile E. superba were higher in total lipid content at the end of the cruise (30

January 2012; 30.7% DW and 34.3% DW, respectively) than at the beginning (6 January 2012,

26.5% DW and 31.8% DW, respectively), but neither of these differences were significant.

DISCUSSION

Species comparison

T. macrura are the smallest, relatively most abundant euphausiids found throughout the W AP

region, concentrated around the northern continental shelf/slope intersection (Nordhausen 1992,

Ward et al. 2004, Ross et al. 2008). While T. macrura are found in stomach contents of Adelie

17

penguins along the WAP, they are usually overlooked as a potentially valuable prey item (Ross et al.

2008). Only three studies to date have discussed T. macrura-penguin dynamics, all in association

with Rockhopper and Macaroni penguins on the sub-Antarctic Heard Island (Brown & Klages 1987,

Klages et al. 1989, Deagle et al. 2007). In one of these, Deagle et al. (2007) found that, euphausiids

constituted 43% of the diet by mass in the late-summer diet of Macaroni penguins, the majority of

which were T. macrura. T. macrura have higher total lipid content than the other krill species

analyzed in this study, and similar to the fish P. antarcacum (although our T. macrura total lipid

values were generally higher than previously reported for this species; Table 3, Falk-Petersen 2000,

Farber-Lorda & Mayzaud 201 0). This could make T. macrura a high quality food item for penguins.

However, T. macrura may not be optimal prey for other reasons. The small size and diffuse

swarming behavior ofT. macrura relative to other krill species (Daly & Macaulay 1988) may make

for inefficient foraging by krill predators. In addition, we found that neutral and polar lipid fractions

in T macrura were distributed evenly, and previous studies have shown that the neutral lipid fraction

is mainly composed of wax esters T. macrura (Hagen & Kattner 1998, Falk-Petersen et al. 2000).

Some penguin species have roughly 20% lower assimilation efficiency for wax esters relative to

triglycerides- the main component of the neutral lipid class for E. superba and P. antarcticum

(Jackson & Place 1990) also potentially detracting from the benefits of a high total lipid content.

E. superba and E. crystallorophias are well-recognized and well-studied prey items for

whales, penguins, and seals (Ross et al. 2008, Chapman et al. 2010, Friedlaender et al. 2011 ). Total

lipid contents for these two species were similar statistically, and they were the only two species

sampled with higher polar than neutral lipid content. Our results are consistent with previous studies

showing that E. superba males have lower total lipid contents and energy densities than females

(Clarke et al. 1980, Farber-Lorda et al. 2009). Unlike previous studies (Farber-Lorda et al. 2009), we

found no significant difference in total lipid between mature female and juvenile E. superba,

although total lipid content is somewhat dependent upon juvenile length (Figure 5, Table 4 ).

18

P. antarcticum had higher total lipid content than E. c1ystallorophias and E. superba, and

higher relative neutral lipid content than any of the three krill species sampled (Table 3),

emphasizing their value as a high energy pr~y item. Previous studies report that the neutral lipid

fraction in P. antarcticum is predominantly composed oftriglycerides, which are assimilated at

higher efficiencies than the wax esters found in T. macrura, E. crystallorophias, and E. antarctica

(Jackson & Place 1990, Whormann et al. 1997, Hagen et al. 2000). P. antarcticum are slow-growing,

notothenioid fish that were historically a consistent prey item for Adelie penguins along the WAP,

but range contractions coincident with regional wanning have now restricted their abundances to

waters south of Adelaide Island (Figure 1) (Emslie & Patterson 2007, Fraser pers. comm.). The

mechanisms associated with this range contraction are unknown, but are likely related to sea-ice

dynamics asP. antarcticum spawning and embryo development takes place below the seasonal pack­

ice (Bottaro et al. 2009). When available, Adelie penguins most frequently take fish 95-120 mm in

standard length, corresponding to year-3 and year-4 age classes (McDaniel & Emslie 2002, Ainley et

al. 2003, Chapman et al. 2011 ). The standard length of fish sampled in the present study ranged from

57.6 - 116.8 mm, with an average of 89.9 mm, representing fish from age classes 2, 3, and 4+

(Hubold & Torno 1989, Chapman eta!. 2011 ). There are few biochemical data available for these

older age classes (Chapman et al. 20 II). We thus added our lipid data to the model of Chapman et

al. (20 11 ), who found that a sigmoidal curve best represented the relationship between P.

antarcticum standard length and total lipid content; our samples fit well within their model (Figure

9), extending their estimations for the year-3 and year-4 age classes.

E. antarctica is a numerically dominant, oceanic species of myctophid mainly found near the

continental shelf/slope interface of the W AP (Donnely & Torres 2008). They also access shelf and

coastal regions through intrusions of warm, UCDW water and are present in both the northern and

southern W AP (Donnely & Torres 2008, Steinberg et al., unpublished data). Their exceptionally high

lipid content, dominated by the neutral lipid fraction, resulted in the highest overall energy densities

19

for all species tested. They are important in the diets of higher predators (Lancraft et al. 2004, Casaux

et al. 2011, HGckstadt et al. 2011 ), and their occurrence in northern, coastal regions of the WAP

makes E. antarctica an energy-rich option for predators residing near Anvers Island, where P.

antarcticum no longer persists (McDaniel & Emslie 2002). However, the neutral lipid fraction of E.

antarctica has been documented as being mainly composed ofwax esters which, as noted above,

may be assimilated at lower efficiency in some species of penguins (Jackson & Place 1990, Phleger

et a!., 1997, Connan et a!. 20 I 0).

For all species analyzed, variations in carbon and nitrogen content were best explained by the

animal's neutral lipid content, in that animals with higher neutral lipid stores had higher carbon and

lower nitrogen content. The neutral lipid fractions of species in this study were likely composed of

triglycerides or wax esters. Triglycerides (found in E. superba and P. antarcticwn) are made from the

attachment of three fatty acids to glycerine through ester bonds and are composed of77% C and 0%

N (Ventura 2006). Wax esters (found in T. macrura, E. c1:ystallorophias, and E. antarctica) are fatty

acids esterified 'vith fatty alcohols and are composed of 81% C and 0% N (Ventura 2006). Proteins,

free amino acids, RNA, DNA, chitin (in krill) and phospholipids contribute to the overall nitrogen

content of an organism (Ventura 2006). As animals retained higher concentrations of neutral lipids,

the overall contribution of the lipids to the dry weight increased, increasing the overall carbon

content relative to nitrogen (Figure 3). The trend is consistent among zooplankton and fish, which

highlights the strong influence of the neutral lipid class on prey elemental composition.

Differences in prey quality between tlte summer of 2009 ami 2010

The majority of our samples were collected in January of 2009 and 2010, years that

represented two archetypal progressions of sea ice for the W AP region. The austral spring of 2008

was marked by early ice retreat over the shelf and inshore, allowing high spring winds to maintain

deeper mixed layers which contributed to the lower Chi a and primary productivity in the summer of

20

2009 (Figure 2c, I 0; Table 1) (Duck! ow et al. 20 I 2b ). Conversely, there was anomalously late sea-ice

retreat in spring 2009. The lingering sea ice protects the waters from high spring winds, while ice

melt enhances stratification contributing to ~hallower mixed layers, likely increasing Chi a and

primary productivity in the summer of 20 I 0 (Figure 2d; Table I) (Vernet et a!. 2008, Duck! ow et a!.

20 I2b ). Total sea-ice extent (defined as total area enclosed by the outer ice edge) in the PAL L TER

sampling grid was I I, I I I km2 in January of 2009 and I2,408 km2 in January of 20 I 0 (S.

Stammerjohn, pers. comm.). The majority of our fish samples were taken in 20 II, which was also a

year of late sea-ice retreat and high Chi a (Table I) (S. Stammerjohn, pers. comm.).

There were significant interannual differences in 2009 vs. 2010 for total length and all prey

quality metrics for T. macrura (both higher in 2009) and E. c1ystallorophias (both higher in 20 I 0).

The prey quality differences are not solely explained by differences in environmental parameters in

2009 vs. 20 I 0, as length is a confounding factor for the analysis. T. macrura lipid content was higher

in 2009 vs. 2010, though this was largely due to the presence of larger individuals in 2009 combined

with a significant positive relationship between length and weight-specific total lipid content for this

species. The same is true for E. crystallorophias, but with the opposite trend. E. clystallorophias

lipid content was higher in 20 I 0 vs. 2009, but this is mainly attributed to the presence of larger

individuals in 2010.

Surprisingly, there were no significant interannual differences in total lipid content of E.

superba in these two opposing sea-ice years. The growth and reproductive cycles of krill are

influenced by environmental parameters such as sea ice, temperature, and food availability (Quetin &

Ross 200 I, Kawaguchi eta!. 2007). Along the W AP, years with early ice retreat (e.g., 2009) are

usually associated with late E. superba spawning, while years with late sea-ice retreat (e.g., 201 0) are

associated with early spawning (Quetin & Ross 200 I). The timing of krill spawning is associated

with the phenology of lipid accumulation during the summer in preparation for winter, so years of

delayed spawning would likely exhibit a delay in lipid accumulation, with the reverse being true in

21

years of early spawning (Chapman et al. 201 0). Chapman et al. (20 1 0) postulated that delayed

spawning and lipid accumulation in years of early ice retreat are due to an associated late seasonal

phytoplankton bloom. Following the Bering Sea model, water column stratification from an early ice /

melt is eroded by spring winds and the bloom is delayed until surface warming from summer

sunlight can restore stratification (Hunt et al. 2002, Chapman et al. 201 0). Chl a peaks later in the

summer, and krill reach maximum lipid accumulation later in the season, with the reverse being true

for years of late sea-ice retreat Our data did not support this hypothesized flexibility in lipid

accumulation phenology. There were no significant interannual differences in total lipid content or

energy densities for E. superba, but juvenile, female, and All E. superba had higher total lipid

contents in 2009 than 2010, and lipid was lower in males during 2009 (Table 1 ). If lipid

accumulation phenology had shifted in response to an early sea-ice retreat and a corresponding delay

in the summer bloom, total lipid contents and energy densities should have been lower for all the

different groups of E. superba sampled in 2009. This suggests that the phenology of E. superba lipid

accumulation in the W AP was not affected by the timing of sea-ice retreat during this study.

P. antarcticum samples had higher total lipid content and energy densities in 2011 than 20 I 0,

which can be attributed to standard length differences in fish collected from each sampling year

(Figure 9).

Regional comparison of prey quality

Demarcation of the North and South sub-regions was based on a 12-year study (1992-2004)

of sea-ice dynamics conducted by Stammerjohn et al. (2008). Waters in the North are characterized

by a highly variable, but lengthy, season of low sea-ice concentration, while the South's sea-ice

season is longer with greater sea-ice persistence (Dierssen et al. 2002, Stammerjohn et al. 2008).

Indeed, summer sea ice was present in the far southern portion of the WAP during every sampling

year of this study. Another feature of the South is a persistent, cyclonic gyre located on the shelf off

22

of Marguerite Bay that has been hypothesized to retain Chi a and larval krill populations (Klinck et

al. 2004, Wiebe et al. 2011 ). Marguerite Bay has also been recognized as a biological hot spot of

enhanced productivity with high Chi a (Vernet et al. 2008), macro- and microzooplankton biomass

(Marrari eta!. 20 II, Price et al., in prep), and apex p{edators (Friedlaender eta!. 20 II).

Significant regional differences in prey quality for the combined 2009 and 20 I 0

juvenile, female, and All E. superba suggest that colder water temperatures and elevated Chi a

increase the prey quality of krill. The physiological response of organisms to climate change is

usually thought of as secondary in importance to broad, ecosystem-based trends, but polar

invertebrates and ectotherms, whose metabolic rates vary with the ambient environment, have narrow

temperature tolerances, which may prove important as the W AP region continues to warm (Peck et al

2004, Duck! ow et al. 20 12). The changing food base in the W AP could also affect E. superba

physiology. Along with the previously discussed reduction in summertime, surface Chi a, there has

been a restructuring of the phytoplankton, where cells >20)lm are more prevalent in the South, and

smaller-celled organisms such as flagellates are becoming more prevalent in the North (Montes­

Hugo et al. 2009).£. sztperba preferentially feed on diatoms and have significantly lower grazing

efficiencies when particles are <20)lm (McClatchie & Boyd 1983, Haberman et al. 2003). Declines

in this preferred food source could also affect the ability of krill to accumulate lipid stores in

preparation for winter. This hypothesis is supported by the significantly higher total lipid content in

the South for juvenile, female, and All E. superba groups (with male E. superba and the fish E.

antarctica exhibiting the same general trend). The overall decline in diatom abundance in the North

also has implications for E. superba winter survival and feeding ecology. Diatoms from the summer

phytoplankton bloom aggregate and sink out of surface waters, providing a strong seasonal pulse of

labile organic carbon to the benthos (Smith et al. 2008). E. superba can access benthic habitats year­

round (including abyssal plains as deep as 3500m) and utilize phytodetritus as an alternate food

source (Clarke & Tyler 2008, Schmidt et al. 2011). The importance, frequency, and seasonality of

23

krill benthic feeding is still not well understood, but decreasing diatom biomass in the north would

equate to a lower POC flux to the benthos, potentially resulting in lower overall food availability for

krill during the winter months. The magnitude of environmental variability between North and South

is likely an important forcing factor affecting regional differences in prey quality. The differences in

upper water column temperature and Chi a between the North and South were more extreme in 20 I 0

than in 2009. This was reflected in North-South prey quality comparisons for each year; there were

no significant North-South differences in 2009 in E. superba prey quality, while juvenile, male, and

All E. superba had significantly higher total lipid content in the South in 2010.

For female E. superba, the increased statistical power associated with the higher number of

samples enabled detection of significant regional differences in the combined 2009 and 20 I 0 data

that were not apparent when each year was analyzed separately. This is most likely related to the

high variability associated with total lipid content for the females as a group. Reproduction in E.

superba peaks during January and February, months coinciding with the seasonal phytoplankton

bloom (Quetin eta!. 1994, Falk-Peterson eta!. 2000). Female krill will produce multiple egg batches

over the spawning season, severely altering their body size and energy stores to accommodate the

brood of developing embryos beneath their carapace (Quetin & Ross 2001 ). This reproductive cycle

has a direct effect on an individual krill's biochemical constituents, as mature gravid females have

higher lipid stored than spent females (Farber-Lorda et a!. 2009).

Regional comparisons of prey quality for both T. macrura and E. crystal!orophias (2009 and

2010 combined) were confounded by interannual variability in length of animals sampled- when

broken down by year, there were no significant differences between North and South for any prey

metric. Thus, comparisons presented in Table 2 cannot be entirely attributed to regional influences.

Additional samples, separated by sex/maturity stage, would be required to detect regional differences

in these species.

24

Relations/tip of prey quality with length, latitude, !tydrograpltic conditions, and time

We found significant relationships between length vs. C, N, and lipid content for T. macrura,

E. clystallorophias, and juvenile E. superba (Table 4, Figure 5), supporting previous studies (Ju &

Harvey 2004, Farber-Lord a & Mayzaud 201 0). The neutral lipid fraction accounted for the majority

of the variability in len[:,rth vs. total lipid content for both T. macrura and E. crystallorophias. This

positive, significant relationship between length and the neutral lipid class has been explained as a

buoyancy regulating mechanism for krill; as their body grows and becomes heavier, relative neutral

lipid content increases so that the animal remains neutrally buoyant (Falk-Petersen 1985). The polar

lipid content was positively, but not significantly, related to length, which is surprising as the polar

fraction is mainly composed of phospholipids which function as structural and storage lipids for

euphausiids (Hagen et al. 1996, Ju et al. 2009). The role of phospholipids as major components of

cell membranes would presumably result in increasing polar lipid concentrations with somatic

growth, but this was not supported by our data.

The environmental and temporal properties included in our analyses (latitude, 0- I 20m mean

water temperature, Julian day sampled, and O-J 20m integrated Chi a) are all highly correlated with

one another, due to the structure of our cruise track (starting in the nmth at Anvers Island and ending

far south at Charcot Island) and the North-South difference in climate regime. Thus, although below

we discuss prey quality in relation to these specific environmental and temporal parameters, we note

it is likely that significant relationships between these properties are the result of a mixture of

influences, instead of just one.

All lipid classes for juvenile, male, and All E. superba were positively related to latitude,

supporting results from our regional analysis (Figure 8). Males and All E. superba were also

significantly, positively related to Chi a, and juveniles and females were significantly, positively

related to Julian date sampled, suggesting that growth trends in juveniles and the reproductive cycle

of females are contributing to the observed latitudinal trend. Although there were significant regional

25

differences in average female E. superba total lipid content, these did not result in a significant

relationship with latitude (although the trend was positive, with a slope of 1.2) or any other

environmental parameters. This is most likely due to the high variability in E. superba females as a

group, associated with the cyclical nature of the spawning season and corresponding changes in

biochemical composition. Chi a explained a larger proportion of the variability in trends in total lipid

content than mean upper water column temperature, and is the environmental parameter contributing

most to observed latitudinal and regional lipid trends.

There was a significant, negative relationship between latitude and Chi a and polar lipid

content, and a significant, positive relationship between mean upper water temperature and polar

lipid content forT. macrura (Table 5). This trend was opposite of that expressed in other species, for

which polar lipid content was generally positively related to latitude and Chi a and negatively related

to mean upper water temperature. (Note- there was no significant relationship between length and

polar lipid content in T. macrura, thus no confounding influence of length differences with latitude).

We suggest that the polar lipid class is more susceptible to environmental variability than the neutral

lipid class for T. macrura. Like most polar zooplankton, it is essential that T. macrura obtain large

deposits of wax esters, the main component of their neutral lipid fraction, during the summer season

to survive through the winter months (Nordhausen 1994, Falk-Petersen et al. 2000). T. macrura is

unusual in that it stores enough lipids to start reproducing in September, before the summer

phytoplankton bloom, (Falk-Petersen et al. 2000, Lee eta!. 2006). Unlike triacylglycerols, wax esters

are used for long-tenn energy storage, and rates of catabolism are highly regulated by lipases (Lee et

al. 2006). T. macrura may thus rely on the polar lipid fraction for energy during the summer, making

this fraction more susceptible to environmental influences (Sargent et a!. 1977, J-Hikanson 1984. Lee

et a!. 2006). Trends in the relationship between environmental parameters and T. macrura prey

quality were opposite to those of other species of krill. Lipid ~ontent in T. macrura was generally

higher in the North, where temperatures were warmer and there was lower Chi a. This is also where

26

maximum abundance of this species is known to occur at the slope/shelf interface (Ross et al. 2008)

thus these conditions must be optimal for T. macrura.

In contrast, all significant relationships between E. c1ystallorophias and environmental or !

!

temporal parameters were associated with the neutral lipid class (but for which length is likely a

confounding variable). The neutral and total lipid fractions for E. c1ystallorophias were significantly,

positively related to latitude and Chi a, and are significantly, negatively related to mean upper water

temperature.

For E. antarctica there was a significant, positive trend in latitude vs. neutral lipid content

and a significant, negative trend in latitude vs. polar lipid content (Table 5). However, there was no

overall significant trend for latitude vs. total lipid content (neutral and polar fractions summed). This

highlights the value of examining lipid classes separately, which can reveal dynamics that are lost

when examining total lipid pools only. The significant, positive relationship between latitude and

neutral lipid content supports the hypothesis that warmer water temperatures and lower Chi a in the

North is associated with lower lipid content. E. antarctica are planktivorous, subsisting mainly on

copepods, amphipods and small euphausiids (Hodde! 1996). Marguerite Bay, where all ofthe South

E. antarctica samples were collected, is considered a biological hotspot characterized by high Chi a

and zooplankton biomass (Ashijan eta!. 2004, Vernet eta!. 2008). The positive relationship between

latitude and neutral lipid content suggests that Marguerite Bay is a better habitat for fish with ready

access to food, which aids in accumulating wax ester stores. The mechanism behind the negative

relationship between latitude (and Julian date sampled) vs. E. antarctica polar lipid content is

unknown, as the function of the polar lipid fraction in E. antarctica, beyond structural purposes, is

poorly understood. Further analysis of individual fatty acids or conducting a compound class

distribution (i.e., measuring gross concentrations of saturated and unsaturated fatty acids, fatty

alcohols, alkanes, etc.) would provide additional insights into these opposing lipid class dynamics.

27

SUMMARY AND CONCLUSIONS

Our characterization of prey quality for five mid-trophic species important to penguin diets

along the W AP further reinforced the importance of fish as a high-energy supplement to a krill-based

diet, due to the relatively higher total lipid content and energy density of fish. The kriii T. macrura is

also high in total lipid content, making it a potentiaily important prey item for apex predators along

the W AP. However, T. macrura 's smaii body size and aggregation behavior may counteract this

benefit, from a predator's perspective.

This study was also the first to relate sea ice dynamics and environmental variability caused

by rapid regional warming with prey quality along the WAP. We found significant regional

differences in E. superba total lipid content and significant relationships between latitude and E.

superba prey quality, as higher total lipid contents were found at higher latitudes. Neutral and polar

lipids equally contributed to total lipid trends, indicating that both are important to krill as they create

energy stores in preparation for winter. Relating E. superba total lipid content to environmental and

temporal variability showed that 0-120 m integrated Chi a and the Julian date that samples were

taken were important drivers in observed latitudinal trends. There was no significant difference in

prey quality for E. superba between January 2009 vs. 2010, two sampling years with different sea icc

dynamics, contradicting the lipid phenology hypothesis presented by Chapman ct al. (20 1 0). For T.

macrura and E. crystallorophias, animal length confounded January 2009 vs. 2010 and North vs.

South comparisons, and more samples separated by length, sex, and maturity stages are required to

accurately characterize any potential trends in prey quality along the sampling grid for these species.

Putting some of our conclusions in the larger context of the W AP ecosystem, rapid regional

warming has fundamentally altered ecosystem structure and, iftrends continue, the northern W AP

will soon be characterized by a Jack of perennial or summer sea ice, warmer water temperatures, and

phytoplankton dominated by smaller cells as opposed to diatoms (Martinson et al. 2008,

28

Stammerjohn eta!. 2008, Montes-Hugo eta!. 2009). We found that these types of conditions are

associated with lower prey quality for E. superba, which could have significant ramifications for the

rich assemblage of endothermic top predators that,make their summer home in the W AP and utilize

krill as their main food source (Croxall et a!. 1992, Chapman eta!. 201 0). Across all sexes and

maturity stages, E. superba in the South had roughly 20% higher total lipid content than in the North,

meaning northern apex predators will have to consume more prey to meet their energetic demands or

increase their reliance on other prey sources. E. superba are also experiencing localized regional

declines in the northern, coastal W AP and there is currently no functional replacement forth is

dominant trophic link (Ross eta!. 2008, Steinberg eta!., unpublished data). The krill T macrura are

high in abundance and prey quality, but their small body size, diffuse aggregation behavior, and high

wax ester content compared to E. superba might make them an inefficient target prey species (Daly

& Macauluy 1988, Jackson & Place 1990, Ross et a!. 2008). The krill E. Cl)JStallorophias and fish P.

antarcticum are also high quality prey, but their ranges are restricted to the South, where summer sea

ice still persists (McDaniel & Emslie 2002, Ducklow et al. 2012a). The myctophid E. antarctica

represents a potential, energy rich option for coastal marine predators residing in the North, although

this fish species' neutral lipid content increased with increasing latitude, also making it a relatively

better quality prey in the South. Incorporating these species-specific prey quality dynamics into

future energetics modeling efforts will increase model accuracy by improving the quality of functions

regulating predator growth.

29

LITERATURE CITED

Ainley DG. 2002. The Adelie penguin: bellwether of climate change. Columbia University Press, New York

Ainley DG, Ballard G, Barton KJ, Karl BJ, Rau GH, Ribic CA, Wilson PR. 2003. Spatial and temporal variation of diet within a presumed metapopulation of Adelie penguins. Condor I 05:95-105

Ainley DG, Ballard G, Blight LK, Ackley S, Emslie SO, Lescroel A, Olmastroni S, Townsend SE, Tynan C, Wilson P, Woehler E. 2009. Impacts of cetaceans on the structure of Southern Ocean food webs. Marine Mammal Science. doi:l O.llll/j.l748-7692.2009.00337.x.

Arrigo KR, Worthen D, Schnell A, Lizotte MP. 1998. Primary production in Southern Ocean waters. Journal ofGeophysical Research-Oceans, 103:15587-15600

Ashijan CJ, Rosenwaks GA, Wiebe PH, Davis CS, Gallager SM, Copley NJ, Lawson GL, Alatalo P. 2004. Distribution ofzooplankton on the continental shelfofMarguerite Bay, Antarctic Peninsula during austral fall and winter, 2001. Deep-Sea Research II, 51 :2073-2098

Atkinson A, Siegel V, Pakhomov E, Rothery P. 2004. Long-term decline in krill stock and increase in salps within the Southern Ocean. Nature, 432:100-103

Atkinson A, Shreeve RS, Hirst A. 2006. Natural growth rates in Antarctic krill (Euphausia superba): II. Predictive models based on food, temperature, body length, sex, and maturity stage. Limnology and Oceanography, 51:973-987

Bernard KS, Steinberg OK, Schofield OME. 2012. Summertime grazing impact ofthe dominant macrozooplankton off the Western Antarctic Peninsula. Deep-Sea Research I, 62:111-122

Bottaro M, Oliveri D, Ghigliotti L, Pisano E, Ferrando S, Vacchi M. 2009. Born among the ice: the first morphological observations of two developmental stages of the Antarctic silverfish Pleuragramma antarcticum, a key species of the Southern Ocean. Reviews in Fish Biology and Fisheries, 19:249-259

Branch TA, Stafford KM, Palacios OM, Allison C, Bannister JL, Burton CLK, Cabrera E, Carlson CA, Galletti Vernazzani B, Gill PC, Hucke-Gaete R, Jenner KCS, Jenner MNM, Matsuoka K, Mikhalev Y A, Miyashita T, Morrice MG, Nishiwaki S, Sturrock V.J, D. Tormosov, Anderson RC, Baker AN, Best PB, Borsa P, Brownell Jr. RL, Childerhouse S, Findlay KP, Gerrodette T, Ilangakoon AD, Joergensen M, Kahn B, Ljungblad OK, Maughan B, McCauley RD, Mckay S, Norris TF, Oman Whale and Dolphin Research Group, Rankin S, Samaran F, Thiele D, Van Waerebeek K, Warneke RM. 2007. Past and present distribution, densities and movements of blue whales Balaenoptera musculus in the Southern Hemisphere and northern Indian Ocean. Mammal Review, 37:116-175

Brekke B, Gabrielsen GW. 1994. Assimilation efficiency of adult kittiwakes and Brunnich guillemots fed capelin and Arctic cod. Polar Biology; 14:279-284

30

Brown CR & Klages NT. I 987. Seasonal and annual variation in diets of Macaroni (Eud;ptes ch7JlSO!ophus ch1ysolophus) and southern Rockhopper (E. chlJlSOcome chlJJSocome) Penguins at Sub-Antarctic Marion Island. Journal of Zoology, 212:7-28

/

Brown M, Kawaguchi S, Candy S, Virtue P. 2010. Temperature effects on the growth and maturation of Antarctic krill (Euphausia superba). Deep-Sea Research II, 57:672-682

Casaux R, Bertolin ML, Carlini A. 2011. Feeding habits ofthree seal species at the Danco Coast, Antarctica: a re-assessment. Polar Biology, 34: I 615-1620

Chapman EW, Hofmann EE, Petterson DL, Fraser WR. 2010. The effects ofvariability in Antarctic krill (Euphausia superba) spawning behavior and sex/maturity stage distribution on Adelie penguin (Pygoscelis adeliae) chick growth: A modeling study. Deep-Sea Research II, 57:543-558

Chapman EW, Hofmann EE, Patterson DL, Ribic CA, Fraser WR. 2011. Marine and terrestrial factors affecting Adelie penguin Pygoscelis adeliae chick growth and recruitment off the western Antarctic Peninsula. Marine Ecology Progress Series, 436:273-289

Clarke A. 1980. The biochemical composition ofkrill, Euphausia superba Dana, from South Georgia. Journal of Experimental Marine Biology and Ecology, 43:221-236

Clarke AH & Tyler PA. 2008. Adult Antarctic krill feeding at abyssal depths. Current Biology, 18:282-285

Clarke J, Manly B, Kerry K, Gardner H, Franchi E, Corsolini S, Focardi S. 1998. Sex differences in Adelie penguin foraging strategies. Polar Biology, 20:248-258

Connan M, Mayzaud P, Duhamel G, Bonnevie BT, Cherel Y. 2010. Fatty acid signature analysis documents the diet of five myctophid fish from the Southern Ocean. Marine Biology, 157:2303-2316

Costa DP & Crocker DE. 1996. Marine mammals ofthe Southern Ocean. In Foundations for ecological research west of the Antarctic Peninsula (ed. Ross RM, Hofmann EE& Quetin LB), pp. 287-301. Washington, DC: American Geophysical Union

Croxall JP. 1992. Southern Ocean Environmental Changes Effects on Seabird Seal and Whale Populations. Philosophical Transactions of the Royal Society of London B Biological Sciences 338: I 19-127

Daly KL & Macaulay MC. 1988. Abundance and distribution ofkrill in the ice edge zone of the Weddell Sea, austral spring 1983. Deep-Sea Research II, 35:21-41

Davis AJ, Jenkinson LS, Lawton JH, Shorrocks B, WoodS. 1998. Making mistakes when predicting shifts in species range in response to global wanning. Nature, 391:783-786.

Deagle BE, Gales NJ, Evans K, Jarman SN, Robinson S, Trebilco R, Hindell MA. 2007. Studying

31

seabird diet through genetic analysis of faeces: A case study on Macaroni Penguins (Eudyptes ch1ysolophus). PLoS ONE 2(9): e831. doi: I 0.13 71 ~journal.pone.0000831

Dierssen HM, Smith RC, Vernet M. 2002. Glacial meltwater dynamics in coastal waters West of the Antarctic Peninsula. Proceedings fn;nn the National Academy of Sciences, 99:1790-1795

/

Dinniman MS, Klinck JM, Hofmann EE. Sensitivity of Circumpolar Deep Water Transport and Ice Shelf Basal Melt along the West Antarctic Peninsula to Changes in the Winds. Journal of Climate, 25(14):4 799-4816

Donnelly J & Torres JJ. 2008. Pelagic fishes in the Marguerite Bay region of the West Antarctic Peninsula continental shelf. Deep-Sea Research II, 55: 523-539

Ducklow HW, Baker K, Martinson DG, Quetin LB, Ross RM, Smith RC, Stammerjohn SE, Vernet M, Fraser W. 2007. Marine ecosystems: The West Antarctic Peninsula. Philosophical Transactions ofthe Royal Society of London B, 362:67-94

Ducklow, H, Clarke A, Dickhut R, Doney SC, Geisz H, Huang K, Martinson DG, Meredith MP, Moeller HV, Montes-Hugo M, Schofield 0, Stammerjohn SE, Steinberg D, Fraser W. 2012a. Ch. 5: The Marine System ofthe Western Antarctic Peninsula. In Antarctic Ecosystems: An Extreme Environment in a Changing World (ed. Rogers AD, Johnston NM, Murphy EJ, Clarke A), pp.121-159. Blackwell, London

Ducklow HW, Schofield 0, Vernet M, Stammerjohn S, Erickson M. 2012b. Multiscale control of bacterial production by phytoplankton dynamics and sea ice along the western Antarctic Peninsula: A regional and decadal investigation. Journal of Marine Systems, 98-99: 26-39

Emslie SO & Patterson WP. 2007. Abrupt recent shift in 13C and 15N values in Adelie penguin eggshell in Antarctica. Proceedings from the National Academy of Science, USA, 104:11666-11669

Falk-Petersen S. 1985. Growth ofthe euphausiids Thysanoessa inermis (Kmyer), Thysanoessa raschii (M. Sars) and Meganyctiphanes norvegica (M. Sars) in a subarctic fjord, North Norway, Canadian Journal of Fisheries and Aquatic Science, 42:14-22

Falk-Petersen S, Hagen W, Kattner G, Clarke A, Sargent J. 2000. Lipids, trophic relationships, and biodiversity in Arctic and Antarctic krill. Canadian Journal of Fisheries and Aquatic Science, 57: 178-191

Farber-Lorda J & Mayzaud P. 20 I 0. Morphology and total lipids in Thysanoessa macrura from the southern part of the Indian Ocean during summer. Spatial and sex differences. Deep-Sea Research II, 57:565-571

Farber-Lorda J, Gaudy R, Mayzaud P. 2009. Elemental composition, biochemical composition and caloric value of Antarctic krill. Implications in Energetics and carbon balances. Journal ofMarine Systems, 78:518-524

Fraser WR & Trivelpiece WZ 1996. Factors controlling the distribution of seabirds: winter-summer heterogeneity in the distribution of Adeiie penguin populations. In Foundations for ecological

32

research west of the Antarctic Peninsula, vol. 70 (ed. Ross RM, Hofmann EE & Quetin LB), pp. 257-272. Washington, DC: American Geophysical Union.

Friedlaender AS, Johnston DW, Fraser WR; Burns J, Halpin PN, Costa DP. 2011. Ecological niche modeling of sympatric krill predatdrs around Marguerite Bay, Western Antarctic Peninsula. Deep-Sea Research II, 58:1729-1740

Golet GH, Kuletz KJ, Roby DD, Irons DB. 2000. Adult prey choice affects chick growth and reproductive success in pigeon guillemots. Auk, 117:82-91

Haberman KL, Ross RM, Quetin LB. 2003. Diet of the Antarctic krill (Euphausia superba Dana) II. Selective grazing on mixed phytoplankton assemblages. Journal of Experimental Marine Biology and Ecology, 283:97-113

Hagen W & Kattner G. 1998. Lipid metabolism ofthe Antarctic euphausiid Thysanoessa macrura and its ecological implications. Limnology and Oceanography, 43:1894-1901

Hagen W, Yen Vleet ES, Kattner G. 1996. Seasonal lipid storage as overwintering strategy of Antarctic krill. Marine Ecology-Progress Series, 134:85-89

Hagen W, Kattner G, Friedrich C. 2000. The lipid compositions ofhigh-Antarctic notothenioid fish species with different life strategies. Polar Biology, 23:785-791

Hakanson JL. 1984. The long and short term feeding conditions in field-caught Calanus paci.ficus, as determined from lipid content. Limnology and Oceanography, 29:794-804

Harrington SA & Thomas PG. 1987. Observations on spawning by Euphausia C1Jlslallorophias from water adjacent to Enderby Land (East Antarctica) and speculation on the early ontogenetic ecology of neritic euphausiids. Polar Biology, 7:93-95

Hoddell R. 1996. Pelagic fish and fish larval distribution in Eastern Antarctic waters (CCAMLR area 58.4.1 ). Honors thesis, University of Tasmania, pp. 180

Hubold G, Tomo AP. 1989. Age and growth of Antarctic silverfish Pleuragramma antarcticum Boulenger, 1902, from the Southern Weddell Sea and Antarctic Peninsula. Polar Biology, 9:205-212

HilcksUidt LA, Burns JM, Koch P, McDonald BI, Crocker DE, Costa DP. 2012. Being a specialist in a changing environment: the diet of the crabeater seal along the western Antarctica Peninsula. Marine Ecology Progress Series, 455: 287-301

Hunt GLJ, Stabeno P, Walters G, Sinclair E, Brodeur RD, Napp JM, Bond NA. 2002. Climate change and control of the southeastern Bering Sea pelagic ecosystem. Deep-Sea Research II, 49:5821-5853

33

Jackson S & Place AR. I 990. Gastrointentinal Transit and Lipid Assimilation Efficiencies in Three Species of Sub-Antarctic Seabird. The Journal of Experimental Zoology, 255:141-154

Ju SJ & Harvey HR. 2004. Lipids as markers of nutritional condition and diet in the Antarctic krill Euphausia superba and Euphausia crystallorophias during austral winter. Deep-Sea Research II, 51 :2199-2214

Ju SJ, Kang HK, KIM WS, Harvey HR. 2009. Comparative lipid dynamics of euphausiids from the Antarctic and Northeast Pacific Oceans. Marine Biology, 156: 1459 - 14 73

Kattner G & Hagen W. 1998. Lipid metabolism ofthe Antarctic euphausiid Euphausia crystallorophias and its ecological implication. Marine Ecology Progress Series, 170:203-213

Kattner G, Hagen W, Lee RF, Campbell R, Deibel D, Falk-Petersen S, Graeve M, Hansen BW, Hirche HJ, J6nasd6ttir SH, Madsen ML, Mayzaud P, Mi.iller-Navarra D, Nichols PO, Paffenhofer G-A, Pond D, Saito H, Sti.ibing D, Virtue P. 2007. Perspectives on marine zooplankton lipids. Canadian Journal of Fisheries and Aquatic Science, 64:1628-1639

Kawaguchi S, Yoshida T, Finley L, Cramp P, NicolS. 2007. The krill maturity cycle: a conceptual model of the seasonal cycle in Antarctic krill. Polar Biology, 30:689-698

Klages NTW, Gales RP, Pemberton D. 1989. Dietary segregation of Macaroni and Rockhopper Penguins at Heard Island. Australian Wildlife Research, 16:599-604

Klinck JM, Hofmann EE, Beardsley RC, Salihoglu B, Howard S. 2004. Water-mass properties and circulation on the west Antarctic Peninsula Continental Shelf in Austral Fall and Winter 2001. Deep-Sea Research II, 51:1925-1946

Lancraft TM, Relsenbichler KR, Robinson BH, Hopkins TL, Torres JJ. 2004. A krill­dominated micronekton and macrozooplankton community in Croker Passage, Antarctica with an estimate offish predation. Deep-Sea Research Part II, 51 :2247-2260

Lee RF, Hagen W, Kattner G. 2006. Lipid storage in marine zooplankton. Marine Ecology Progress Series, 307:273-306

Loeb V, Siegel V, Holm-Hansen 0, Hewitt R, Fraser W, Trivelpiece S. 1997. Effects of sea­ice extent and krill or salp dominance on the Antarctic food web. Nature, 387:897-900

Makarov RR & Denys CJ. 1981. Stages of sexual maturity of Euphausia superba Dana. BIOMASS Handbook 11:1-11

Marrari M, Daly KL, Timonin A, Semenova T. 2011. The zooplankton ofMarguerite Bay, Western Antarctic Peninsula- Part I: Abundance, distribution, and population response to variability in environmental conditions. Deep-Sea Research II, 58:1599-1613

34

Martinson DG & McKee DC. 2012. Transport of warm upper circumpolar deep water onto the western Antarctic Peninsula Continental Shelf, Ocean Sciences. 8, 433-442, doi: 1 0.5194/os-8-433-2012

Martinson DG, Stammerjohn SE, Iannuzzi P~A, Smith RC, Vernet M. 2008. Western Antarctic Peninsula physical oceanography and spatia-temporal variability. Deep-Sea Research 11, 55:1964-1987

Mauchline 1.1970. Measurement of Body Length of Euphausia superba Dana. BIOMASS Handbook No.4. SCAR/SCOR/IABO/ACMRR, pp. 1-9

Mayzaud P, Boutoute M, Alonzo F. 2003. Lipid Composistion ofthe euphausiids Euphausia vallentini and 111ysanoess macrura during summer in the Southern Indian Ocean. Antarctic Science, 15:463-47

McClatchie S & Boyd CM. 1983. Morphological study of sieve efficiencies and mandibular surfaces in the Antarctic krill, Euphausia superba. Canadian Journal of Fisheries and Aquatic Science, 40:955-967

McClintock, J, Ducklow, H, and W. Fraser. 2008. Ecological responses to climate change on the Antarctic Peninsula. American Scientist, 96:302-310

McDaniel JD & Emslie SO. 2002. Fluctuations in Adelie penguin prey size in the mid to late Holocene, north Marguerite Bay, Antarctic Peninsula. Polar Biology, 25:618-623

Moline MA, Karnovsky NJ, Brown Z, Divoky GJ, Frazer TK, Jacoby CA, Torrese JJ, Fraser WR. 2008. High latitude changes in ice dynamics and their impact on polar marine ecosystems. Annals of the New York Academy of Sciences, 1134:267-319

Montes-Hugo M, Doney SC, Ducklow HW, Fraser W, Martinson D, Stammerjohn SE, Schofield 0. 2009. Recent changes in phytoplankton communities associated with rapid regional climate change along the western Antarctic Peninsula. Science, 323: 1470-1473

Nagy KA & Obst BS. 1992. Food and energy-requirements of Adelie penguins (Pygoscelis adeliae) on the Antarctic Peninsula. Physiological Zoology, 65, 1271-1284

Nordhausen W. 1992. Distribution and growth of larval and adult Thysanoessa macrura (Euphausiacea) in the Bransfield Strait region, Antarctica. Marine Ecology Progress Series 83, 185-196.

Nordhausen W. 1994. Distribution and vertical migration of the euphausiid 17J.ysanoessa macrura Gerlache Strait, Antarctica. Marine Ecology Progress Series, 83:185-196

Pakhomov EA, Froneman PW, Perissinotto R. 2002. Salp/krill interactions in the Southem Ocean: spatial segregation and implications for the carbon flux. Deep-Sea Research 11, 49:1881-1907

Peck LS, Webb KE, Bailey OM. 2004. Extreme sensitivity ofbiological function to temperature in Antarctic marine species. Functional Ecology, 18:625- 630

35

Phleger CF, Nichols PO, Virtue P. I 997. The lipid, fatty acid and fatty alcohol composition of the myctophid fish Electrona antarctica: high level of wax esters and food-chain implications. Antarctic Science, 9:258-265

Poerschmann J & Carlson R. 2006. New fractionation scheme for lipid classes based on "in-cell fractionation" using sequential pressurized liquid extraction. Journal of Chromatography, I 127: I 8-25

Quetin LB & Ross RM. 200 I. Environmental variability and its impact on the reproductive cycle of Antarctic krill. American Zoologist, 41 :74-89

Quetin LB, Ross RM, Clarke A. 1994. Krill energetics: seasonal and environmental aspects of the physiology of Euphausia superba. In Southern Ocean Ecology: the BIOMASS Perspective (ed. El-Sayed SZ), pp. 165-184. Cambridge University Press, Cambridge

Ricker WE & Westrheim SJ. 1979. Bias in age-length keys-reply. Journal of the Fisheries Rsearch Board of Canada, 36(9): 1 I 60- I 16 I

Ross RM, Quetin LB, Martinson DG, Iannuzzi RJ, Stammetjohn S, Smith RC. 2008. Palmer L TER: Patterns of Distribution of Five Dominant Zooplankton Species in the Epipelagic Zone West ofthe Antarctic Peninsula, I993- 2004. Deep Sea Research II, 55:2086-2105

Sargent JR, Gatten RR, Corner EDS, Kilvington CC. 1977. On the nutrition and metabolism of zooplankton. XI. Lipids in Calanus helgolandicus grazing Biddulphia sinensis. Journal of the Marine Biological Association ofthe United Kingdom, 57:525:533

Schmidt K, Atkinson A, Steigenberger S, Fielding S, Lindsay MCM, Pond OW, Tarling GA, Klevjer TA, Allen CS, NicolS, Achterberg EP. 2011. Seabed foraging by Antarctic krill: Implications for stock assessment, bentho-pelagic coupling, and the vertical transfer of iron. Limnology and Oceanography, 65:1411-1428

Smith CR, Mincks S, DeMaster D. 2008. The FOODBANCS project: Introduction and sinking fluxes of organic carbon, chlorophyll-a and phytodetritus on the western Antarctic Peninsula continental shelf. Deep-Sea Research II, 55:2404-2414

Smith RC & Stammerjohn SE. 2001. Variations of surface air temperature and sea ice extent in the Western Antarctic Peninsula (WAP) region. Annals of Glaciology, 33:493-500

Smith RC, Fraser WR, Stammerjohn SE. 2003. Climate variability and ecological response of the marine ecosystem in the westem Antarctic Peninsula (WAP) region. In Climate variability and ecosystem response at long- term ecological research sites (ed. Greenland 0, Goodin DG, Smith RC), pp. I 58-173. New York, NY: Oxford University Press

Stammerjohn SE, Martinson DG, Smith RC, Iannuzzi RA. 2008. Sea Ice in the Westem Antarctic Peninsula Region: Spatia-Temporal Variability from Ecological and Climate Change Perspectives. Deep Sea Research II, 55:2041-2058

36

Stammerjohn SW, Masson R, Rind D, Martinson D. 2012. Regions of rapid sea ice change: An inter­hemispheric seasonal comparison. Geophysical Research Letters 39, L06501

Steinberg OK, Martinson DG, Costa DP. Two decades of pelagic ecology of the Western Antarctic Peninsula. Oceanography, in press

Van de Waal DB, Verschoor AM, Verspagen JMH, van Donk E, Huisman J. 2009. Climate­driven changes in the ecological stoichiometry of aquatic ecosystems. Frontiers in Ecology and the Environment. 8: 145-152

Vaughan DG, Marshall GJ, Connolley WM, Parkinson C, Mulvaney R, Hodgson DA, King .IC, Pudsey CJ, Turner J. 2003. Recent rapid regional climate wanning on the Antarctic Peninsula. Climatic Change, 60:243-74

Ventura M. 2006. Linking biochemical and elemental composition in freshwater and marine crustacean zooplankton. Marine Ecology Progress Series, 327:233-246

Vernet M, Martinson D, Iannuzzi R, Stammerjohn S, Kozlowski W, Sines K, Smith R, Garibotti I. 2008 Primary production within the sea-ice zone west of the Antarctic Peninsula: I -Sea ice, summer mixed layer, and irradiance. Deep Sea Research II, 55:2068-208

Ward P, GrantS, Brandon M, Siegel V, Sushin V, Loeb V. Griffiths H. 2004. Mesozooplankton community structure in the Scotia Sea during the CCAMLR 200 survey: January-February 2000. Deep-Sea Research II, 51:1351-1367

Wiebe PH, Ashjian CJ, Lawson GL, Pifiones A, Copley NJ. 2011. Horizontal and veritical distribution of euphausiid species on the Western Antarctic Peninsula U.S. Globec Southern Ocean study site. Deep-Sea Research II 58:1630-1651

Williams TO. 1995. The penguins: Spheniscidae. Oxford University Press, New York, NY

Whitehouse MJ, Atkinson A, Wars P, Korb RE, Rothery P, Fielding S. 2009. Role of krill versus

bottom-up factors in controlling phytoplankton biomass in the northern Antarctic waters of

South Georgia. Marine Ecology Progress Series, 393:69-82

Wohrmann APA, Hagen W, Kunzmann A. 1997. Adaptations ofthe Antarctic silverfish

PleuraJ::,>ramma antarcticum (Pisces: Nototheniidae) to pelagic life in high-Antarctic waters.

Marine Ecology Progress Series, 151 :205-218

37

Table 1. January 2009 vs. 2010 comparison of environmental parameters and prey quality. Averages and ranges (in parentheses) for properties measured in January 2009, 2010, and 2011 for prey species: Euphausiids- T.macrura, Thysanoessa macrura; E. c1ystallorophias, Euphausia c1ystallorophias; Juvenile E. superba, Juvenile Euphausia superba; Male E. superba, Male Euphausia superba; Female E. superba, Female Euphausia superba;All E. superba, All Euphausia superba; Fish - P. antarct;cum, Pleuragramma antarcticum; E. antarctica, Electrona antarctica. P­values are given for all significant results (where p:::: 0.05); ns, not significant; dash (-) indicates data only spanned one sampling year and no interannual comparisons were made. All prey metric data are presented as dry weight (DW), and energy density as kilojoules (kJ).

38

ProQertv 2009 2010 2011 Significance 0-120m Average Water Temp (°C) -0.01 (-0.98-0.72) -0.58 (-1.69-0.36) p < 0.001 0-120m Integrated Chi a ( mg m·2) 52.19 (17.47-103.59) 119.42 (25.38-623.43) 209.16 (10.78-621.70) p < 0.001 ('09 vs '10)

p < 0.001 ('09 VS '11) ns ('10 vs '11)

T. macrura Nitrogen (mg N mg-1 DW-1) 0.080 (0.067-0.118) 0.089 (0.082-0.097) p = 0.003 T. macrura Carbon (mg C mg-1 DW-1) 0.46 (0.39-0.52) 0.42 (0.40-0.47) p = 0.03 T. macrura (mg C:mg N) 5.88 (3.78-7.58) 4.79 (4.28 -5.13) p = 0.003 T. macrura Total Lipids (%DW) 54.20 (42.78-71.31) 31.23 (20.59-41.43) p < 0.001 T. macrura Total Length (mm) 14.22 (10.14-17.39) 12.30 (10.62-13.68) p = 0.03

E. crystallorophias Nitrogen (mg N mg-1 DW-1) 0.099 (0.094-0.10) 0.088 (0.081-0.093) p < 0.001 E. crystallorophias Carbon (mg C mg·1 DW-1) 0.39 (0.37-0.41) 0.43 (0.42-0.46) p < 0.001 E crystallorophias (mg C:mg N) 3.92 (3.67-4.29) 4.94 (4.66-5.16) p < 0.001 E. crystallorophias Total Lipids (%DW) 21.02 (16.95-27.01) 34.54 (28.18-42.10) p = 0.009 E. crystallorophias Total Length (mm) 29.78 (26.83-32.19) 33.66 (30.20-36.38) p = 0.005

juvenile E. superba Nitrogen (mg N mg-1 DW-1) 0.084 (0.077-0.10) 0.086 (0.072-0.105) liS

juvenile E. superba Carbon (mg C mg-1 DW-1) 0.43 (0.40-0.47) 0.42 (0.36-0.45) p = 0.03 juvenile E. superba (mg C:mg N) 5.14 (4.36-6.02) 4.82 (3.92-5.70) ns juvenile E. superba Total Lipids (%DW) 36.88 (19.61-64.51) 3 2.2 7 (14.5 7 -46.68) liS

juvenile E. superba Total Length (mm) 33.66 (27.01-38.87) 31.34 (28.08-38.05) p = 0.03

Male E. superba Nitrogen (mg N mg-1 DW-1) 0.096 (0.076-0.106) 0.090 (0.071-0.106) liS Male E. superba Carbon (mg C mg-1 DW-1) 0.41 (0.36-0.46) 0.40 (0.32-0.50) liS

Male E. superba (mg C:mg N) 4.34 (3.50-5.59) 4.45 (3.37-6.39) ns Male E. superba Total Lipids (%DW) 25.79 (14.06-45.06) 28.23 (17.36-54.27) liS

Male E. superba Total Length ( mm) 46.96 (40.84-50.18) 49.28 (43.23-52.62) p = 0.02 Male E. superba Energy Density (kj g·1 DW-1) 18.92 (16.02-23.27) 19.77 (16.75-24.36) liS

Female E. superba Nitrogen (mg N mg·1 DW-1) 0.088 (0.076-0.097) 0.088 (0.078-0.104) liS

Female E. superba Carbon (mg C mg-1 DW-1) 0.43 (0.39-0.47) 0.42 (0.37-0.46) liS

Female E. superba (mg C:mg N) 4.96 (4.27-6.21) 4.86 (3.68-5.70) liS

Female E. superba Total Lipids (%DW) 36.42 (29.20-54.72) 32.20 (12.36-42.10) ns Female E. superba Total Length (mm) 47.68 (40.34-51.88) 48.86 (43.41-52.60) liS

Female E. superba Energy Density (kj g-1 DW-1) 22.43 (19.68-25.33) 21.48 (19.00-23.74) liS

All E. superba Nitrogen (mg N mg·1 DW-1) 0.089 (0.076-0.106) 0.088 (0.071-0.106) liS All E. superba Carbon (mg C mg·1 DW-1) 0.43 (0.36-0.47) 0.41 (0.32-0.50) p = 0.003 All E. superba (mg C:mg N) 4.88 (3.50-6.21) 4.69 (3.37-6.39) liS

All E. superba Total Lipids (%DW) 34.04 (14.06-64.51) 30.74 (12.36-54.27) liS All E. superba Total Length (mm) 42.07 (27.01-51.88) 43.96 (28.08-52.62) ns All E. superba Energy Density (k) g DW-1) 21.6 (16.02-25.33) 20.58 (16.75-24.36) liS

P. antarcticum Nitrogen (mg N mg-1 DW-1) 0.09 0.090 (0.084-0.097) 0.082 (0.068-0.098) liS ('10 VS '11) P. antarcticum Carbon (mg C mg-1 DW-1) 0.46 0.4-7 (0.46-0.50) 0.50 (0.44-0.54) liS ('10 VS '11)

\, P. antarcticum (mg C:mg N) 5.16 5.29 (4.72-5.90) 6.14 (4.97-6.97) p = 0.05 ('10 VS '11) P. antarcticum Total Lipids (%DW) 42.56 34.52 (27.41-38.08) 41.96 (37.98-46.82) p = 0.02 ('10 VS '11) P. antarcticum Standard Length (mm) 57.62 83.13 (70.35-88.83) 99.40 (85.40-116.82) p = 0.01 ('10 vs '11) P. antarcticum Energy Density (kj g·1 DW·1) 23.35 (22.36-24.77) 25.18 (23.74-26.93) p = 0.04 ('10 VS '11)

E. antactica Nitrogen (mg N mg-1 DW-1) 0.059 (0.056-0.064) E. antactica Carbon (mg C mg-1 DW-1) 0.61 (0.56-0.64) E. antactica (mg C:mg N) 10.30 (9.24-11.4 7) E. antactica Total Lipids (%DW) 54.96 ( 44.41-59.82) E. antactica Standard Length (mm) 76.48 (54.49-90.09)

_E. antactica Energy Densi!J: (k! g·1 DW-1) 31.92 (29.96-32.68)

39

Table 2. Regional comparison of environmental parameters and prey quality. Regional averages and ranges (in parentheses) for properties measured in the North and South -as defined in Figure 1, for prey species: Euphausiids- Tmacrura, Thysanoessa macrura; E. CIJ!Stallorophias, Euphausia c1ystallorophias; Juvenile E. superba, Juvenile Euphausia superba; Male E. superba, Male Euphausia superba; Female E. superba, Female Euphausia superba; Ail E. superba, Ail Euphausia superba; Fish- P. antarcticum, Pleuragramma antarcticum; E. antarctica, Electrona antarctica. Ail regional comparisons are combined 2009 and 201 0 data, except for 2011 Chi a, P. antarcticum, and E. antarctica comparisons, which included 20 II data. P-values are given for all significant results (where p:S0.05); ns, not significant; dash (-) indicates data only spanned one region and no comparisons were made. All prey metric data are presented as dry weight (OW), and energy density as kilojoules (kJ).

40

ProQer~ North South Significance 0-120m Average Water Temp (0 C) 0.12 (-0.60-0.72) -0.57 (-1.69-0.21) p < 0.001 2009 & 2010, 0-120m Integrated Chi a (mg m·Z) 60.78 (20.45-169.91) 103.82 (17.4 7 -623.43) p = 0.05 2011, 0-120m-Integrated Chi a (mg m·Z) 39.92 (10.78-73.70) 255.84 (24.41-621.70) p < 0.001

T. macrura Nitrogen (mg N mg·l DW·l) 0.087 (0.075-0.118) 0.079 (0.067-0.093) ns T. macrura Carbon (mg C mg·l DW·l) 0.43 (0.39-0.51) 0.46 (0.42-0.52) p = 0.03 T. macrura (mg C:mg N) 5.07 (3.78-6.25) 5.97 (4.56-7.58) p = 0.04 T. macrura Total Lipids (%DW) 46.22 (20.59-71.31) 43.81 (22.89-70.40) ns T. macrura Total Length (mm) 12.56 (10.14-14.87) 14.68 (11.62-17.39) p = 0.01

E. crystallorophias Nitrogen (mg N mg·l DWI 0.100 (0.096-0.104) 0.090 (0.081-0.100) ns E. crystallorophias Carbon (mg C mg·l DWl) 0.38 (0.37-0.39) 0.42 (0.38-0.46) ns E. crystal/orophias (mg C:mg N) 3.77 (3.67-3.94) 4.72 (3.91-5.16) p = 0.009 E. crystal/orophias Total Lipids (%DW) 19.01 (16.96-22.41) 31.66 (20.54-42.10) p < 0.001 E. crystal/orophias Total Length (mm) 28.17 (26.83-29.4 7) 33.09 (30.20-36.38) p = 0.003

juvenile E. superba Nitrogen (mg N mg·l DW-1) 0.083 (0.074-0.094) 0.086 (0.072-0.106) ns

juvenile E. superba Carbon (mg C mg·l DW·l) 0.41 (0.36-0.45) 0.42 (0.39-0.47) ns juvenile E. superba (mg C:mg N) 4.91 (3.94-5.72) 4.99 (3.92-6.02) ns juvenile E. superba Total Lipids (%DW) 30.96 (14.57-48.89) 37.57 (21.64-64.51) p = 0.007 juvenile E. superba Total Length (mm) 32.66 (27.01-38.87) 32.32 (28.08-37.17) ns

Male E. superba Nitrogen (mg N mg·l DW-1) 0.093 (0.071-0.11) 0.091 (0.071-0.10) ns

Male E. superba Carbon (mg C mg·l DW-1) 0.39 ( 0.32-0.46) 0.42 (0.36-0.50) ns Male E. superba (mg C:mg N) 4.20 (3.37-5.59) 4.63 (3.50-6.39) ns Male E. superba Total Lipids (%DW) 24.55 (14.06-45.06) 30.54 (15.03-54.27) ns Male E. superba Total Length (mm) 48.63 ( 40.84-52.62) 48.17 ( 43.23-52.53) ns Male E. superba Energy Density (kj g·l DW-1) 19.23 (16.02-23.27) 19.84 (16.38-24.36) ns

Female E. superba Nitrogen (mg N mg·l DW-1) 0.092 (0.083-0.104) 0.086 (0.076-0.097) p = 0.005 Female E. superba Carbon (mg C mg-1 DW·l) 0.42 (0.37 -0.45) 0.43 (0.38-0.47) ns Female E. superba (mg C:mg N) 4.60 (3.68-5.09) 5.06 (4.05-6.21) p = 0.01 Female E. superba Total Lipids (%OW) 30.77 (12.46-39.50) 36.26 (25.47-54.72) p = 0.03 Female E. superba Total Length (mm) 48.18 ( 40.34-51.88) 48.31 (43.41-52.60) ns

Female E. superba Energy Density (kj g·l DWI) 21.83 (19.00-23.75) 22.13 (19.68-25.33) ns

All E. superbo Nitrogen (mg N mg·l DW-1) 0.091 (0.071-0.11) 0.087 (0.071-0.106) ns All E. superba Carbon (mg C mg·l DW·l) 0.41 (0.32-0.46) 0.43 (0.37-0.50) p = 0.001 All E. superba (mg C:mg N) 4.55 (3.37-5.72) 4.92 (3.50-6.37) p = 0.004 All E. superba Total Lipids (01(JDW) 28.51 (12.36-48.89) 35.21 (15.03-64.51) p < 0.001 All E. superba Total Length (mm) 42.61 (27.01-52.62) 42.51 (28.08-52.60) ns

All E. superba Energy Density (kj g·l OW·!) 20.46 (16.02-23.75) 21.45 (16.38-25.33) ns

P. antarcticum Nitrogen (mg N mg·l DWl) 0.085 (0.068-0.098) P. antarcticum Carbon (mg C mg·l OW· I) 0.49 (0.44-0.54) P. antarcticum (mg C:mg N) 5.73 ( 4.72-6.97) P. antarcticum Total Lipids (%DW) 38.60 (27.41-46.82) P. antarcticum Standard Length (mm) 89.93 (57.62-116.82) P. antarcticum Energy Density (kj g·l DW·l) 24.63 (22.36-26.93)

E. antactica Nitrogen (mg N mg·l DW·l) 0.062 (0.061-0.064) 0.058 (0.056·0.062) p = 0.05 E. antactica Carbon (mg C mg·l DW-1) 0.59 (0.59-0.60) 0.61 (0.56-0.65) ns E. antactica (mg C:mg N) 9.52 (9.24-9.80) 10.50 (9.49-11.47) ns

E. antactica Total Lipids (%DW) 50.60 (44.41-56.79) 56.05 (50.74-59.82) ns

E. antactica Standard Length (mm) 78.87 (77.64-80.10) 75.88 (54.49-90.09) ns

E. antactica Energy Densi!Y: (k! g·l OWl) 31.94 (31.82-32.07) 31.92 (29.96-32.68) ns

41

Table 3. Prey quality measurements by species. Averages and ranges (in parentheses) for all sampling years combined (2009-2011) for prey species: Euphausiids- T.macrura, Thysanoessa macrura; E. Cl~vstallorophias, Euphausia cJystallorophias; Juvenile E. superba. Juvenile Euphausia superba: Male E. superba, Male Euphausia superba; Female E. superba, Female Euphausia superba; All E. superba, All Euphausia superba; Fish- P. antarcticum. Pleuragramma antarcticum: E. antarctica. Electrona antarctica. All euphausiid length data are given as total length, while fish data are given as standard length. All prey metric data are presented as dry weight (DW), and energy density as kilojoules (kJ).

Property T. macrura £ crystallorophias Juvenile£ superba Male£ superba Female E. superba All E. superba P. antarcticum E. antarctica

Nitrogen (rug N mg·1 DW-1) 0.083 (0.067-0.118) 0.092 (0.081-0.104) 0.085 (0.072-0.106) 0.092 (0.071-0.106) 0.088 (0.076-0.104) 0.088 ( 0.071-0.106) 0.085 (0.068-0.098) 0.059 (0.056-0.064) Carbon (rug C mg·1 DW-1) 0.45 (0.39-0.52) 0.42 (0.37-0.46) 0.42 (0.36-0.47) 0.40 (0.32-0.50) 0.43 (0.37-0.47) 0.42 (0.32-0.50) 0.49 (0.44-0.54) 0.61 (0.56-0.64) Mg C:mg N 5.48 (3.78-7.58) 4.53 (3.67-5.16) 4.96 (3.92-6.02) 4.42 (3.37-6.39) 4.91 (3.68-6.22) 4.77 (3.37-6.39) 5.73 (4.72-6.97) 10.30 (9.24-11.4 7) Neutral Lipids (%DW) 22.47 (11.25-36.35) 11.47 (4.62-19.30) 14.60 (5.75-29.54) 10.03 (1.85-25.72) 13.65 (2.54-20.44) 12.85 (1.85-29.54) 25.71 (15.03-35.53) 48.42 (33.74-53.97) Polar Lipids (%DW) 22.35 (9.19-35.29) 17.67 (11.93-28.46) 19.74 (7.00-46.42) 17.34 (9.30-30.09) 20.78 (9.81-34.28) 19.34 (7.00-46.42) 12.90 (7.10-20.03) 6.54 (4.54-10.68) Total Lipids (%DW) 45.02 (20.59-71.31) 29.13 (16.95-42.10) 34.52 (14.57-64.51) 27.37 (14.06-54.27) 34.43 (12.36-54.72) 32.26 (12.36-64.51) 38.60 (27.41-46.82) 54.96 (44.41-59.82) Total/Standard Length (mm) 13.52 (10.14-17.39) 32.10 (26.83-36.38) 32.47 (27.01-38.87) 48.41 (40.84-52.62) 48.27 (40.3-52.60) 42.55 (27.01-52.62) 89.92 (57.62-116.32) 76.48 (54.49-90.09) Energy Density fkl g·1 DW-1) 19.52 (16.02-24.36) 22.04 (19.00-25.33) 21.07 (16.02-25.33) 24.63 (22.36-26.93) 31.92 (29.96-32.68)

42

Table 4. Relationship between length vs. elemental composition and lipid fraction by species. Average individual length (mm) vs. elemental composition (mg N mg-1 ow-1 ormg C mg-1 OW-1

) and lipid fractions(% dry weight. OW) for euphausiid prey species: T.macrura, Thysanoessa macrura: E. Cl}'Stallorophias. Euphausia ClJ'Stallorophias; Juvenile E. superba, Juvenile Euphausia superba. There were no significant length- composition relationships for other species (not shown). Regression equations are: Elemental Composition (mg N mg-1 ow-1 ormg C mg-1 OW-1

) or Lipid Fraction (%OW)= m (Average Individual Length. mm) +b. R2 values are given in front of p-values for all significant relationships (where p::::; 0.05); ns. not significant. See also Figure 4.

S ecies Elemental Com osition b) RZ Li id Fraction Slo e (m) Interce t b) RZ T. macrura Nitrogen (mg N mg·l DW-1) 0.49 (p < 0.001) Neutral 2.87 -17.18 0.52 (p = 0.0003)

Carbon (mg C mg-1 DW-1) 0.35 (p = 0.006) Polar p = 0.15, ns C:N 0.57 (p < 0.001) Total 4.14 -12.12 0.31 (p=0.01)

E. crystallorophias Nitrogen (mg N mg·l DW·l) 0.15 0.65 (p < 0.001) Neutral 1.13 -24.74 0.45 (p = 0.006)

Carbon (mg C mg-1 DW-1) 0.006 0.21 0.44 (p = 0.007) Polar p = 0.081, ns C:N 0.15 -0.4 0.64 (p < 0.001) Total 1.94 -33.09 0.48 (p = 0.004)

juvenile E. superba Nitrogen (mg N mg-1 DW-1) -0.001 0.12 0.21 (p = 0.02) Neutral 0.75 -9.91 0.20 (p = 0.005)

Carbon (mg C mg-1 DW-1) p = 0.08, ns Polar 0.8 -6.26 0.13 (p = 0.03) C:N 0.09 2.02 0.28 (p = 0.004) Total 1.46 -12.97 0.21 (n = 0.004)

43

/

n 21 22 21

15

15 15

39

39 39

Table 4. Relationship between length vs. elemental composition and lipid fraction by species. Average individual length (mm) vs. elemental composition (rng N mg-1 D\V-1 ormg C mg-1 DW-1

) and lipid fractions(% dry weight. DW) for euphausiid prey species: T.macrura, Thysanoessa macrura: E. cT:vstallorophias, Euphausia cTystallorophias; Juvenile E. superba, Juvenile Euphausia superba. There were no significant length -composition relationships for other species (not shown). Regression equations are: Elemental Composition (mg N mg-1 DW-1 ormg C mg-1 DW-1

) or Lipid Fraction (%DW) = m (Average Individual Length, mm) +b. R2 values are given in front of p-values for all significant relationships (where p:::; 0.05); ns. not significant. See also Figure 4.

S ecies Elemental Com osition b) R2 Li id Fraction Slo e (m) In terce t (b) RZ T. macrura Nitrogen (mg N mg-1 DW-1) 0.49 (p < 0.001) Neutral 2.87 -17.18 0.52 (p = 0.0003)

Carbon (mg C mg-1 DW-1) 0.35 (p = 0.006) Polar p = 0.15, ns C:N 0.57 (p < 0.001) Total 4.14 -12.12 0.31 (p = 0.01)

E. crystallorophias Nitrogen (mg N mg-1 DW-1) -0.002 0.15 0.65 (p < 0.001) Neutral 1.13 -24.74 0.45 (p = 0.006) Carbon (mg C mg-1 DW-1) 0.006 0.21 0.44 (p = 0.007) Polar p = 0.081, ns C:N 0.15 -0.4 0.64 (p < 0.001) Total 1.94 -33.09 0.48 (p = 0.004)

Juvenile E. superba Nitrogen (mg N mg-1 DW-1) -0.001 0.12 0.21 (p = 0.02) Neutral 0.75 -9.91 0.20 (p = 0.005) Carbon (mg C mg-1 DW-1) p = 0.08, ns Polar 0.8 -6.26 0.13 (p = 0.03) C:N 0.09 2.02 0.28 ro = 0.004) Total 1.46 -12.97 0.21 ro = 0.004)

43

/

n 21 22 21

15

15 15

39

39 39

Table 5. Relationship between environmental parameters and lipid composition by species. Various regional or environmental parameters vs. lipid fractions(% dry weight. DW) for prey species: Euphausiids- Tmacrura, Thysanoessa macrura; E. crystallorophias. Euphausia Cl}'Stallorophias; Juvenile E. superba. Juvenile Euphausia superba; Male E. superba. Male Euphausia superba; Female E. superba, Female Euphausia superba; All E. superba, All Euphausia superba: Fish- P. antarcticum. Pleuragramma antarcticum: E. antarctica. Electrona antarctica. Regression equations are: Lipid Fraction (%DW) = m (Environmental Parameter)+ b. Regressions for data from all sampling years combined (2009-20 I I), unless otherwise noted. R2 are given in front of p-values for all significant relationships (where p:S0.05); ns. not significant. See also Figure 7.

44

/

Latitude (DO) 0-120m Mean Water Temp ("C) 0-120m Integrated Chi a (mg m·2)

Slope Species Lipid Fraction Slope (m) Intercept (b) R> (m) Intercept (b) R> Slope (m) Intercept (b) R> n

T. mocrura Neutral p = 0.75, ns p = 0.14, ns -0.26 36.32 0.72 (p =0.0003) 21 Polar -3.15 230.35 0.24 (p = 0.03} 10.43 21.97 0.51 (p = 0.0003) -0.27 36.66 0.71 (p = 0.0003) zz Total p = 0.17, ns 16.96 44.97 0.39 (p = 0.003) -0.53 73.32 0.79 (p = <0.0001) 21

E. crystalforophias Neutral 2.04 -128.78 0.39 (p = 0.01) -5.12 8.22 0.67 (p = 0.0002) 0.03 6.67 0.67 (p = 0.006) 15 Polar p = 0.19, ns p = 0.07, ns p = 0.08, ns 15 Total 3.25 -194.19 0.35 (p = 0.02) -8.29 23.87 0.63 (p = 0.0004) 0.05 20.83 0.70 (p =0.004) 15

Juvenile, Neutral 1.32 -73.43 0.24 (p = 0.02) p = 0.77, ns p = 0.16, ns 39 E. superba Polar 2.31 -134.45 0.24 (p = 0.002) p = 0.55, ns p = 0.09, ns 39

Total 3.50 -198.69 0.26 (p = 0.001) p = 0.87, ns p = 0.07. ns 39

Male, E. superba Neutral 1.60 -95.61 0.17 (p = 0.02) p = 0.23, ns p = 0.25, ns 34 Polar 1.41 -76.07 0.17 (p = 0.02) p = 0.17, ns 0.06 12.79 0.19 (p = 0.01) 34 Total 3.01 -171.69 0.23 (p = 0.004) p = 0.14, ns 0.10 20.39 0.13 (p = 0.04) 34

Female, Neutral p = 0.08, ns p = 0.42, ns p = 0.16, ns 36 E. superba Polar p = 0.35, ns p = 0.66, ns p = 0.66, ns 36

Total p = 0.12, ns p = 0.47, ns p = 0.68, ns 36

All E. superba Neutral 1.35 -77.18 0.14 (p = <0.0001) p = 0.73, ns p =0.12, ns 109 Polar 1.52 -81.70 0.15 (p = <0.0001) p = 0.27, ns 0.40 16.68 0.06 (p = 0.01) 109 Total 2.83 -156.06 0.20 (p = <0.0001) p = 0.42, ns 0.05 28.22 0.05 (p = 0.02) 109

P. antarcticum Neutral p = 0.35, ns *2009&2010 data only p = 0.98, ns p = 0.63, ns 13 Polar p = 0.42, ns p = 0.50, ns p = 0.29, ns 13 Total p = 0.53, ns p = 0.64, ns p = 0.90, ns 13

E. antarctica Neutral 3.32 -174.41 0.44 (p = 0.04) *2011 data not available p = 0.07, ns 10 Polar - 1.34 96.19 0.58 (p = 0.01) p = 0.31, ns 10 Total p = 0.14, ns p = 0.07, ns 10

45

A)

62•s

Year Sampled , All E. superba

e 2009 and 2010

* 2009

7o•s 7s•w

sa•w

120 Krlomeler;

sa•w B)

s2•s sa•w s2·s~ '

Year Sampled, All E. superba Ill

• 2009 E. crysla11oroph ias

• 2009 T. macrura

• 2010 E crysla11orophias

• 2010 T. macrura

X 2009 P. an tarclicum

+ 2010 P. an larcl icum

* 2011 P. antarc ticum

©

, ©

________ ,_..----

7o•s 75' W

' ' ~ ·"'_) ! ~ ( - ,

<Jl . ~ \ -...:.:,. t --· ~ \....·~

ss•w

~ (:).< u ·o u ~ >¢ d.

120 Kr lomeler;

Figure 1. PAL L TER study region and stations sampled. Map of study region where A) E. superba. Euphausia superba and B) E. crystallorophias, Euphausia crystallorophias: T. macrura, Thysanoessa macrura; P. antarcticum, Plueragramma antarcticum: E. antarctica. Electrona antarctica. were collected. An: Anvers Island, Ad: Adelaide Island, MB: Marguerite Bay, Ch: Charcot Island. The North and South regions used for comparative analysis are indicated in brackets. and based on contrasting sea-ice characteristics described by Stamme1john et al. (2008). Shades of grey are bathymetry, with light grey representing the continental shelf and dark grey the continental slope and abyssal plain. The continental shelf is roughly 200 km wide and averages 430 min depth (Ducklow et al. 201 2a).

46

68' W

A) 68' W

C) 70' S

S!il'.lans 1'1'1!h Temp OJ!il

B) 75•w · • Sto11005 w11h Cht.1 d(l tn

D)

47

120m Integrated Ch la mglm'2 __j 25.-30 0

__] 301- 400

62' 5-__j 40 1-50 0 50 1 - 75 0

68'W

' 68 ' W

68'W

Figure 2. Regional maps of mean upper water column temperature and integrated Chi a. Mean water temperature 0-120, (°C, A and B) and O-l20m integrated Chi a (mg m-2

, C and D) for the study region from CTD and discrete bottle data collected during the 2009 (A and C) and 20 l 0 (B and D) summer cruises in January. Stations occupied are solid, black circles.

48

0.11 0.7

o o 0 0 0 00 0

~ 0.10 0@ 0

~

~ T T

~ ~t~ 0.6 OJ 0 E 0.09

OJ E z 0

~ o.J Ol .., .§. 0 0 .. T 1:: 0.08 T 0

2 c c 2 0 c 0 0 0 c 0.07 0 0 0 Q) T c Ol .8 0.4 _g

ro z 0.06 0 go 0

0 0.05 0.3

0 10 20 30 40 0 10 20 30 40

A) Neutral Lipid Content (%DVV) B) Neutral Lipid Content (%DW)

0.11 . 0.7

• ~ 0.10 •• ~ .. ~

0 ~ • 0.6 1 "' OJ 0

E 0.09 OJ

z E Ol 0 I ../ "' .§. OJ

..... 0.08 • 0 .§. c ..... 0.5 Q) c

~ 0 c Q)

0 .... 0 c 0 () 0

c 0.07 0 0

Q) c Q)

OJ 0 0 E .D 0.4 -~ ro z 0.06 "' 0

0.05 0.3

0 10 20 30 40 50 60 0 10 20 30 40 50 60

C) Neutral Lipid Content (%DW) D) Neutral Lipid Content (%DVV)

49

Figure 3. Relationship between neutral lipid content and elemental composition by species. Neutral lipid content(% dry weight, OW) vs. elemental composition (mg N mg-1 ow-1 ormg C mg-1 ow-1

) for prey species. A and B represent species whose neutral lipids were mainly composed oftriglycerides- (O) All Euphausia superba; ( v) P!euragramma antarcticum. C and D represent species whose neutral lipids were mainly composed of wax ester- ( o) Thysanoessa macrura; ( o) Euphausia crystallorophias: ( v) Electrona antarctica. A linear regression was the best fit for data in graphs: A, Band D. Regression equations are (Elemental Composition. mg N mg-1 ow-1 ormg C mg- 1

0\V-1) = m(Neutral Lipid Content, %0\V) +b. A quadratic polynomial was the best fit for data in graph C. Regression equation was:

(Nitrogen Content mgN mg-1 OW-1) = a(Neutral Lipid Content. %DWi + m(Neutral Lipid Content, %0\V) +b. Regressions are for

combined 2009 & 20 I 0 data and all are significant (p:S0.05).

so

60 l I

so I l I

~ 40 ~ Q

~ I rn I ::: 30 1 0.

:::3 -13 0

E- 20 ~

i

10 -1 I

I 0

• 2009 Total Lipids 11 2010 Total Lipids 2011 Total Lipids

*

T. macrura E. crystallorophias

Juvenile E. superba

Male E. superba Female E. superba

Prey Species

*

All E. superba P. antarcticum E. antarctica

Figure 4. January 2009 vs. 2010 comparison of total lipid content by species. Average total lipid content (%dry we ight. OW) in 2009, 20 I 0. and 20 I L across the entire sampling grid , for prey species: Euphausi ids - T macrura, Thysanoessa macrura: E. CI)'Sfallorophhts, Euphausia CI)'Sfallorophias: Juvenile E. superba, Juvenile Euphausia superba: Male E. superba. Male Euphausia superba: Female E. superba. Female Euphausia superba: All E. superba, All Euphausia superba. Fish- P. antarcticum, Pleuragramma antarcticum: E. antarctica. Electrona antarctica. Error bars are l standard error. Asterisks(*) indicate significant differences in tota l lipid content between years. See Table l for test statistics.

51

80

70

~ 60

0 ~50 Ul

~ :.:J 40 ro 0

r- 30 I 0

20

10 10 12

A)

• •

0

0 0

•

14

T. macrura

•

•• •

16

Average Individual Total Length (mm)

80

70

~ 60

0

~50 Ul

X ::J 40 ro 0 1-

30

20

10 20

C)

18

50

~40 0 :;R 0 -(/) "0 30 ·a. ::i ro 0 1- 20

E. crystalforophias

0

0 00

0 ~0 • /

• •• • •

10+---------~r----------r----------,-----------,

20 20 25 30 35 40

B) Average Individual Total Length (mm)

Juvenile E. superba

0

0 0 oo

0 0 •

oo

~ • • 0

0 0 0 0o 0

0 0 0

30 25 35

Average lndividua!Total Length (mm)

52

0

40

0 = 2009 0 = 2010

Figure 5. Relationships between length and total lipid content by species. Average individual total length (mm) vs. total lipid content (% dry weight, DW) for the euphausiids: T macrura. Thysanoessa macrura; E. crystallorophias, Euphausia crystallorophias: Juvenile E. superba, Juvenile Euphausia superba. Regressions are for combined 2009 & 2010 data. See Table 4 for regression equations and statistics.

53

60 l

J ~40 0 ;;g?_ 0 '-' VI

:::: 30 c. ~

~ ~ 20

10

0

:<.'/)­:<.'>­

'?1-(j

"'-.~

. '?1-c, ::v"

:<.o« ~0

c}'/)-

6~ ~-

*

;-o'?l­~:<.

c,'>-~ ~·

·~~ ~

.:;.~

~

• North I" South

;-o'?l-q}

c,'>-~ ~·

~::r;.~

*

x} q}

c,'>-~ ~-

~~ :$-'/)-

"'~ Prey Species

*

;-o'?l­~:<.

c,~~ ~~·

't'

'>-.t$' :O(j

:<.(j .__'?:-

'/)-~ ~ ·

.c.,i'b-.(}

~"""' '/)-~ ~-

Figure 6. Regional comparison of total lipid contents by species. Average total lipid content (as% dry weight. OW) in the North and the South regions. for all sampling years (2009-20 11 ). fo r prey species: Euphausiids - T. macrura, Thysanoessa macrura: E. crystallorophias, Euphausia CI)"Stallorvphias: Juvenile E. superba. Juvenile Euphausia superba; Male E. superba. Male Euphausia superba: Female E. superba. Female Euphausia superba: All E. superba. All Euphausia superba. Fish- P antarcticum. Pleuragramma antarcticum; E. antarctica. Electrona antarct ica. Error bars are 1 standard error. Asterisks(*) indicate significant differences in total lipid content between regions. See Table 2 for test statistics.

54

.... (J)

·c -o .... -(])

-o 0 u

A)

-64.5

-65.5

0 -66.5

Q <!) u ~ 5 -67.5

-68.5

-69 .5

0

c:i

0

0

0

0

0

0

"' U)

9 C>

';- ';-

0

•

0 0

I

•

0 ••

•

•

"' 0

•

0-120m Average Water Tem pe rature (•C)

0

B)

-64.5 I

-65.5

0 -66.5 0 <!) "0 2 rn ...J -67.5

-68.5

-69 .5

0

Oo.

•

0

0

10 20

Julian Date Sampled

eo 0

0

30

C)

-64.5

-65.5 f 0

0 -66.5 Q <!) "0 ::>

-~

5 -67 .5

-61c\.5

-69 .5

•

•

0

0

0

100

0

• = 2009 0 = 2010

0

200 300

0-lZOm In tegrated Chi a (mg m·')

Figure 7. Relationship between latitude and environmental or temporal parameters. Latitude (decimal degrees, DD) vs.0-120m mean water temperature, Julian day. and 0-120m integrated Chi a. X- and Y -axes in plots have been reversed to better depict North to South trend. Regressions are for 2009 and 2010 combined data. Regression equations are (Property)= m(Latitude) + band are significant (p < 0.05); see Table 5 for regression equations and statistics.

55

~

l ~ "' '\' w E ..... (\)

~ ~ I

I Ill

3 ~ · ... -:. '

"' _J

~ I <0 '\' ·;::::

-o ..... -~ I 0>

"? 0 u I __

0 ":"

0

A)

All E. superba

.·fl

••• •

• •

•

20

..

-40

Total Lipids by %DW

• •

~

"' '\'

~

Ill

" :::; " z 'I' ro _J

<0 'I'

• 0> "?

0 ":"

60 0

B)

Juvenile E. superba

•• • • •

.. •

•

-20 40

Total Li pids by %DW

•

60

C)

~ r

I c'J

:t I

Ill ""0 :::; " ·~ "? I _J

<0 '\' .

~n

I 0 ":"

0

Male E. superba

• 1,\ - • -

~- •

• • I e •

•

• • \-. • • •

\ • 10 20 30 40 so

Total Lipids by %DW

Figure 8. Relationship between latitude and total lipid content for Euphausia superba. Latitude (decimal degrees, DO) vs . All , juvenile, and male E. superba. Euphausia superba, total lipid content(% dry weight. OW). X- and Y -axes in plots have been reversed to better depict North to South trend. Regressions are for 2009 and 20 I 0 combined data. Regression equations are (Total Lipid Content) = m(Latitude) + b. Regressions are significant (p < 0.05): see Table 5 for regression formulas and statistics.

56

60

50

0 45

0 Hubold & Hagen 1997 ~ 40 0 ~ e.... 35 ...... c (j)

0 0 <> Hagen et al. 2007

0

D This study, 2009 c

30 0 0 fa This study, 2010 ""0 :e.. 25 ....J D This study, 2011 ro 0 20 f-

15

10 0 20 40 60 80 100 120 140

Fish Standard Length (mm)

Figure 9. Relationship between length and total lipid content relationship for Pleuragramma antarcticum. Individual standard length (mm) vs. total lipid content(% dry \Veight, OW) for the fish Pleuragramma antarcticum. Figure adapted from Chapman et al. (2011). The formula for the P!euragramma antarcticum energy density function is: Lip(L) = [(LiPmax-LiPmin)* 1/(1 +e-klip(L-l..lipSOl)]+Lipmin· Where Lip(L) = lipid content of fish with total length L (mm): LiPmax =maximum lipid content (as proportion dry mass); LiPmin =minimum lipid content (proportion dry mass): klip =controls the rate of increase in lipid content with increasing L (set to O.lmm-1

): and k1irso = L at which the lipid content is 50% of its maximum value (set to 50mm).

57

APPENDIX

Table Al. Relationship between environmental and temporal parameters. Regression statistics for the relationship between latitude (decimal degrees, DO), Julian date, 0-120m mean water temperature (0C) and 0-120m integrated Chi a (mg m-2

). Regression equations are in linear form: y = m(x) +b. R2 values are given in front p-values for all significant relationships (where p :S 0.05); ns. not significant. Regressions are for combined 2009 and 20 I 0 data. See also Figure 6.

2009 & 2010

Water Temp Date 0-120m Integrated Chi a (mg m-2)

2 · - z-- 2 R = 0.47, p < 0.0001 R = 0.46, p = 0.0001 R = 0.20, p = 0.018

Latitude {DD) m = (-0.24), b = (15.82) m = (2.60), b = (-157.81) m = (15.69), b = (-964.54)

R2 = 0.17, p = 0.034 R2 = 0.20, p = 0.019

0-120m Mean Water Temp rCJ m = (-4.43), b = (14.35) m = (-0.005), b = (0.05)

p = 0.16, ns Julian Date Sampled m = (0.03), b = (13.21)

58

Table A2. Julian date vs. lipid composition by species. Relationship between Julian date vs. lipid fractions(% dry weight, DW) for prey species: Euphausiids- Tmacrura, Thysanoessa macrura; E. crystallorophias, Euphausia CI)JStallorophias; Juvenile E. superba, Juvenile Euphausia superba; Male E. superba, Male Euphausia superba; Female E. superba, Female Euphausia superba; All E. superba, All Euphausia superba. Fish- P. antarcticum, Pleuragramma antarcticum; E. antarctica, Electrona antarctica. Regression equations are: Lipid Fraction (%DW) = m (Julian Date)+ b. R2

given in front of p-values for all significant relationships (where p:SO.OS); ns, not significant.

S~ecies Li~id Fraction (%DW) Sioee (m} Interce~t (b} RZ n

T. macrura Neutral p = 0.18, ns 21

Polar p = 0.36, ns 22

Total p = 0.84, ns 21

E. crystallorophias Neutral 0.46 1.46 0.28 (p = 0.04) 15

Polar p = 0.43, ns 15

Total p = 0.09, ns 15

Juvenile E. superba Neutral p = 0.06, ns 39

Polar 0.52 12.3 0.18 (p = 0.007) 39

Total 0.77 23.63 0.19 (p = 0.006) 39

Male E. superba Neutral p = 0.58, ns 34

Polar p = 0.64, ns 34

Total p = 0.93, ns 34

Female E. superba Neutral 0.33 8.76 0.21 (p = 0.005) 36

Polar p=0.16, ns 36

Total 0.56 26.07 0.17 (p = 0.01) 36

All E. superba Neutral p = 0.10, ns 109

Polar p = 0.11, ns 109

Total p = 0.07, ns 109

P. antarcticum Neutral p = 0.07, ns 13

Polar p = 0.37, ns 13

Total p = 0.09, ns 13

E. antarctica Neutral p = 0.22, ns 10

Polar -0.43 13.85 0.61 (p = 0.008) 10

Total = 0.60, ns 10

59

VITA

KATE ELIZABETH RUCK

Born in Annapolis, Maryland on August 13, 1985. Graduated from Hickory High School in 2003. Earned a B.S. in Biology with a minor in Mathematics from James Madison University in 2007. Worked as a research intern with Dr. William Peterson at Hatfield Marine Science Center of Oregon State University examining growth rates of the krill Euphausia pac(fica through cohort analysis. Worked as a lab technician for Dr. Fu-Lin Chu at the Virginia Institute of Marine Science, College of William & Mary, investigating the modification of fatty acids and sterols by planktonic heterotrophic protists in the marine food web. Worked as a field technician under Dr. George Kling, from University of Michigan, for the Arctic Long-Term Ecological Research project (ARC-LTER). Entered the Masters program at the Virginia Institute of Marine Science, College of William & Mary, under graduate advisor Dr. Deborah K. Steinberg in 2009.